Yao, Bo (2011) Mental simulations in comprehension of direct versus
indirect speech quotations.
PhD thesis.
http://theses.gla.ac.uk/3067/
Copyright and moral rights for this thesis are retained by the author
A copy can be downloaded for personal non-commercial research or
study, without prior permission or charge
This thesis cannot be reproduced or quoted extensively from without first
obtaining permission in writing from the Author
The content must not be changed in any way or sold commercially in any
format or medium without the formal permission of the Author
When referring to this work, full bibliographic details including the
author, title, awarding institution and date of the thesis must be given
Glasgow Theses Service
http://theses.gla.ac.uk/
[email protected]
Mental Simulations in Comprehension of
Direct versus Indirect Speech Quotations
Bo Yao, B.Sc.
Submitted in fulfilment of the requirements for the Degree of Ph.D.
School of Psychology
College of Science and Engineering
University of Glasgow
December 2011
Acknowledgements
I would like to express my greatest gratitude to my supervisors Dr Christoph Scheepers
and Prof Pascal Belin for their invaluable guidance and support throughout the time of my
Ph.D. research.
Special thanks to Dr Christoph Scheepers who took me on board despite my Biology
background. Your saying, „man with motivation can achieve anything‟ has inspired me to
accomplish my Ph.D. adventure, and has encouraged me to pursue an independent
academic career.
I am also grateful to Anna Murray who helped with stimulus preparation and data
collection for some part of this research, to Rebecca Watson for her kind tutorials on fMRI
data analysis, and to Dr Sara Sereno and Prof Paddy O‟Donnell for their kind support and
inspiration.
Finally, I would like to thank my parents for their unconditional love and support, and my
uncle for guiding me to be a confident scientist and a respectable person.
Page | 2
Author’s declaration
I declare that this thesis is my own work carried out under normal terms of supervision.
Page | 3
Abstract
In human communication, direct speech (e.g., Mary said: „I‟m hungry‟) coincides with
vivid paralinguistic demonstrations of the reported speech acts whereas indirect speech
(e.g., Mary said [that] she was hungry) provides mere descriptions of what was said.
Hence, direct speech is usually more vivid and perceptually engaging than indirect speech.
This thesis explores how this vividness distinction between the two reporting styles
underlies language comprehension. Using functional magnetic resonance imaging (fMRI),
we found that in both silent reading and listening, direct speech elicited higher brain
activity in the voice-selective areas of the auditory cortex than indirect speech, consistent
with the intuition of an „inner voice‟ experience during comprehension of direct speech. In
the follow-up behavioural investigations, we demonstrated that this „inner voice‟
experience could be characterised in terms of modulations of speaking rate, reflected in
both behavioural articulation (oral reading) and eye-movement patterns (silent reading).
Moreover, we observed context-concordant modulations of pitch and loudness in oral
reading but not straightforwardly in silent reading.
Finally, we obtained preliminary
results which show that in addition to reported speakers‟ voices, their facial expressions
may also be encoded in silent reading of direct speech but not indirect speech.
The results show that individuals are more likely to mentally simulate or imagine reported
speakers‟ voices and perhaps also their facial expressions during comprehension of direct
as opposed to indirect speech, indicating a more vivid representation of the former. The
findings are in line with the demonstration hypothesis of direct speech (Clark & Gerrig,
1990) and the embodied theories of language comprehension (e.g., Barsalou, 1999; Zwaan,
2004), suggesting that sensory experiences with pragmatically distinct reporting styles
underlie language comprehension.
Page | 4
Publications obtained during the course of the study, of which, materials (or part) will be
presented in the thesis:
Journal papers:



Yao, B., Belin, P., & Scheepers, C. (2011). Silent reading of direct versus indirect
speech activates voice-selective areas in the auditory cortex. Journal of Cognitive
Neuroscience, 23(10), 3146-3152. Presented in Chapter 2.
Yao, B., & Scheepers, C. (2011). Contextual modulation of reading rate for direct
versus indirect speech quotations. Cognition, 121(3), 447-453. Presented in Chapter 6.
Yao, B., Belin, P., & Scheepers, C. (under review). Brain „talks over‟ boring quotes:
Top-down activation of voice-selective areas while listening to monotonous direct
speech quotations. NeuroImage. Presented in Chapter 3.
Conference and workshop presentations:








Yao, B., Belin, P., & Scheepers, C. (June 2011). Activation of voice-selective brain
areas in response to written and auditory quotations. Poster presentation at 17th Annual
Meeting of the Organization on Human Brain Mapping (HBM), Quebec City, Canada.
Yao, B., Belin, P., & Scheepers, C. (April 2011). Poster presentation at Activation of
voice-selective brain areas in response to written and auditory quotations. International
Workshop on Voice and Speech Processing in Social Interations (VSPSI), Glasgow,
UK.
Yao, B., & Scheepers, C. (April 2011). Spontaneous voice acting during reading aloud
of direct versus indirect speech. Poster presentation at International Workshop on
Voice and Speech Processing in Social Interations (VSPSI), Glasgow, UK.
Yao, B., Belin, P., & Scheepers, C. (March 2011). Activation of voice-selective brain
areas in response to written and auditory quotations. Poster presentation at 24th Annual
Conference on Human Sentence Processing (CUNY), Stanford, USA.
Yao, B., & Scheepers, C. (March 2011). Spontaneous voice acting during reading
aloud of direct versus indirect speech. Poster presentation at 24th Annual Conference
on Human Sentence Processing (CUNY), Stanford, USA.
Yao, B., & Scheepers, C. (March 2011). Contextual modulation of reading rate in oral
and silent reading of direct versus indirect speech. Poster presentation at 24th Annual
Conference on Human Sentence Processing (CUNY), Stanford, USA.
Yao, B., Belin, P., & Scheepers, C. (January 2011). Activation of voice-selective brain
areas in response to written and auditory quotations. Poster presentation at Language
in Glasgow (LanG), Glasgow, UK.
Yao, B., Belin, P., & Scheepers, C. (September 2010). Top-down activation of voice
areas in the auditory cortex during silent reading of direct vs. indirect speech. Selected
talk at 16th Annual Conference on Architectures and Mechanisms for Language
Processing (AMLaP), York, UK.
Page | 5
Table of Contents
Abstract ..................................................................................................................................4
Table of Contents ...................................................................................................................6
Chapter 1
1.1
Thesis introduction .........................................................................................11
Direct and indirect speech and their pragmatic functions .....................................11
1.2
The vividness distinction between direct and indirect speech in verbal
communication .................................................................................................................13
1.3
The representational consequences of the vividness distinction in language
comprehension ..................................................................................................................15
1.4
Thesis hypothesis and motivations ........................................................................19
1.5
Thesis overview .....................................................................................................23
Part 1: Differential brain activity in voice-selective areas during comprehension of
direct versus indirect speech .........................................................................................24
Part 2: The acoustic aspects of the „inner voice‟ in comprehension of direct versus
indirect speech ..............................................................................................................25
Part 3: The visual aspect of the direct-speech representations: Mental simulations of
facial expressions? ........................................................................................................28
PART 1: Differential Brain Activity in Voice-Selective Areas during Comprehension of
Direct versus Indirect Speech...............................................................................................30
Chapter 2
Brain „hears‟ voices when reading direct quotations: Silent reading of direct
versus indirect speech activates the human voice-selective areas (Experiment 1) ..............31
2.1
Introduction ...........................................................................................................31
2.2
Method...................................................................................................................33
2.2.1
Participants .....................................................................................................33
2.2.2
Stimuli and design ..........................................................................................33
2.2.3
Procedure........................................................................................................34
Page | 6
2.2.4
MRI acquisition..............................................................................................36
2.2.5
Data analysis ..................................................................................................37
2.3
Results ...................................................................................................................38
2.4
Supplemental analyses ..........................................................................................39
2.5
Discussion .............................................................................................................42
Chapter 3
Brain „talks over‟ boring quotes: Top-down activations of voice-selective
areas when listening to monotonous direct speech utterances (Experiment 2) ....................46
3.1
Introduction ...........................................................................................................46
3.2
Method...................................................................................................................48
3.2.1
Participants .....................................................................................................48
3.2.2
Stimuli ............................................................................................................48
3.2.3
Procedure........................................................................................................49
3.2.4
MRI acquisition..............................................................................................50
3.2.5
Data analysis ..................................................................................................50
3.3
Results and discussion ...........................................................................................55
3.3.1
Whole brain and ROI analyses .......................................................................55
3.3.2
Parametric modulation analyses .....................................................................57
3.3.3
Comparing brain activation patterns between studies ....................................64
3.4
Summary and general discussion ..........................................................................65
PART 2: The Acoustic Aspects of the „Inner Voice‟ in Comprehension of Direct versus
Indirect Speech .....................................................................................................................69
Chapter 4
Contextual modulation of pitch representations in oral and silent reading of
direct versus indirect speech ................................................................................................70
4.1
Introduction ...........................................................................................................70
Page | 7
4.2
Experiment 3 .........................................................................................................77
4.2.1
Participants .....................................................................................................77
4.2.2
Stimuli and design ..........................................................................................77
4.2.3
Procedure........................................................................................................79
4.2.4
Data analyses ..................................................................................................79
4.2.5
Results and discussion ...................................................................................81
4.3
Experiments 4, 5 and 6 ..........................................................................................88
4.3.1
Participants .....................................................................................................88
4.3.2
Stimuli and design ..........................................................................................88
4.3.3
Procedure........................................................................................................91
4.3.4
Results and discussion ...................................................................................92
4.4
Summary and general discussion ..........................................................................96
Chapter 5
Contextual modulations of loudness representations in oral and silent reading
of direct versus indirect speech ............................................................................................99
5.1
Introduction ...........................................................................................................99
5.2
Experiment 7 .......................................................................................................101
5.2.1
Participants ...................................................................................................101
5.2.2
Stimuli and design ........................................................................................101
5.2.3
Procedure......................................................................................................102
5.2.4
Results and discussion .................................................................................102
5.3
Experiment 8 .......................................................................................................104
5.3.1
Participants ...................................................................................................104
Page | 8
5.3.2
Stimuli and design ........................................................................................104
5.3.3
Procedure......................................................................................................106
5.3.4
Results and discussion .................................................................................107
5.4
General discussion ...............................................................................................110
Chapter 6
quotations
Contextual modulation of reading rates for direct versus indirect speech
......................................................................................................................112
6.1
Introduction .........................................................................................................112
6.2
Experiment 9 .......................................................................................................113
6.2.1
Participants ...................................................................................................113
6.2.2
Stimuli and design ........................................................................................113
6.2.3
Procedure......................................................................................................114
6.2.4
Results and discussion .................................................................................115
6.3
Experiment 10 .....................................................................................................116
6.3.1
Participants ...................................................................................................116
6.3.2
Stimuli and design ........................................................................................116
6.3.3
Procedure......................................................................................................116
6.3.4
Results and discussion .................................................................................117
6.4
Summary and general discussion ........................................................................121
PART 3: The Visual Aspect of the Direct-Speech Representations: Mental Simulations of
Facial Expressions? ............................................................................................................124
Chapter 7
Are facial expressions also encoded in silent reading of direct speech?
(Experiment 11) .................................................................................................................125
7.1
Introduction .........................................................................................................125
7.2
Participants ..........................................................................................................128
Page | 9
7.3
Stimuli and design ...............................................................................................128
7.4
Procedure .............................................................................................................130
7.5
Results and discussion .........................................................................................131
Chapter 8
Conclusions and closing remarks .................................................................136
Appendices .........................................................................................................................141
References ..........................................................................................................................195
Page | 10
Chapter 1 Thesis introduction
1.1
Direct and indirect speech and their pragmatic functions
Direct speech (e.g., Mary said: „I‟m hungry‟) and indirect speech (e.g., Mary said [that]
she was hungry) are two important reporting styles in human communication (Coulmas,
1986). Direct speech is commonly used when the reported utterance is framed in the
reported speaker‟s words (e.g., Mary‟s words) whereas indirect speech is applied when the
reported utterance is paraphrased in the reporter‟s words (e.g., my words).
Direct speech and indirect speech have distinct pragmatic functions. The more traditional
view (e.g., Leech & Short, 1981; Thomson & Martinet, 1986) argues that direct speech is a
verbatim quotation of the reported utterance whereas indirect speech is a paraphrased
description of the reported utterance. While such a hypothesis sufficiently interprets the
usage of indirect speech, it appears to be inadequate in explaining certain phenomena of
direct speech usage.
First of all, a verbatim hypothesis implies that a direct speech quotation can only be made
upon an existing original utterance.
However, there are cases in which no original
utterance has ever existed to be quoted. For example, people‟s thoughts (e.g., Mary
thought: „I‟m hungry‟) have never been uttered into words; therefore the content of
thoughts can only be inferred rather than be „quoted‟ verbatim. Moreover, utterances in
fiction do not exist in the real world and it would be illogical to quote Gandalf (i.e., a
wizard in the Lord of the Rings) word-for-word.
Secondly, a verbatim hypothesis assumes that the quoted original utterance has to be a
linguistic statement uttered by a human being such that it can be quoted word-for-word.
But in some cases, the original utterances are non-linguistic vocalisations which cannot be
understood without imaginary interpretations. For instance, My kitten was like: „Meow.. I
want milk, Mummy!‟ is believed to be a direct quotation of a cat‟s meowing. Since kittens
do not speak English, there could have been no linguistic information available to be
quoted. Thus, the „I want milk‟ quotation can only be an inferred or imagined linguistic
interpretation rather than a verbatim quotation.
Page | 11
Finally, a verbatim hypothesis also overestimates our memory capacity. Most of the time,
we simply cannot remember every single word that has been said. Empirical research
shows that it is surprisingly difficult for participants to recall an utterance word-for-word
even after just a few seconds (e.g., Anderson, 1974; Hjelmquist & Gidlund, 1985; Jarvella,
1971; Lehrer, 1989; Sachs, 1967; Stafford, Burggraf, & Sharkey, 1987). Moreover, certain
expressive elements allowed in direct speech such as „so-and-so‟, „blah-blah-blah‟ and
„this and that and everything else‟ clearly suggest missing out of the original words. For
example, consider the following sentence Then we come back to me in the studio and I say,
„The film you have just seen was an attempt to blah blah blah.‟ (British National Corpus,
G12, 731). In this context, „blah blah blah‟ was not the exact words the reported speaker
said. Instead, it was intended to represent „some cliché comments about a film‟ which the
reporter – who was in fact the reported speaker himself – simply could not remember or
did not even bother to report word-for-word. Such an intended „unfaithfulness‟ in direct
speech reporting is clearly incongruent with the verbatim hypothesis.
In contrast, a recent proposal by Clark and Gerrig (1990) adequately accommodates the
linguistic phenomena that have been discussed above, namely, direct quotations of „neverexisted‟, non-linguistic and omitted/forgotten utterances. They argue that direct speech,
instead of being a verbatim quotation of the reported utterance1, provides non-serious and
selective demonstrations of the reported speech acts. For example, when the reporter says
Mary said: „I‟m hungry!‟, he/she means that the reported speaker Mary is hungry. In this
sense, „I‟m hungry!‟ is not a serious statement that the reporter meaning him/herself is
hungry but a demonstration of the reported speech utterance. Moreover, the reporter may
selectively demonstrate certain aspects of the original utterance such as dramatizing her
vocal features while omitting what she actually said. As demonstration, direct speech does
not have to be a verbatim quotation of the reported utterance; in some circumstances, it can
simply be „constructed‟ regardless of the original utterance‟s existence or nature (Clark &
Gerrig, 1990; Tannen, 1986, 1989). Thus, even though the original utterances have never
existed, are non-linguistic or are omitted or forgotten, they can still be „constructed‟ to
1
Note: the demonstration account is not „anti-verbatim‟; it actually acknowledges that at least in certain
institutional settings, such as in newspapers, law courts or academic contexts, verbatim quotations are
conventionally encouraged (cf. Clark & Gerrig, 1990; Short, Semino, & Wynne, 2002).
Page | 12
depict what the reported situations may have been like or what the reported speakers may
have said.
More importantly, Clark and Gerrig‟s proposal recognises that in speech reporting,
paralinguistic information such as speakers‟ voices, facial expressions and body languages
conveys important pragmatic meanings which linguistic information alone cannot
effectively communicate. As a result, the core of Clark and Gerrig‟s (1990) proposal lies
in its emphasis on the paralinguistic forms (i.e., how they are reported) rather than the
linguistic content (i.e., what was reported) of direct speech reporting. They argue that the
key distinction between direct speech and indirect speech lies in the stylistic features of the
two reporting styles: Direct speech is an important stylistic device for enlivening stories; it
provides vivid demonstrations of the reported speech acts, enabling listeners to experience
what it would be like to see, hear or feel what the reported speaker did. By contrast,
indirect speech provides mere descriptions of what was said, without depicting the
paralinguistic information that was communicated in the reported speech acts.
This
pragmatic distinction predicts that direct speech is perceptually more vivid and engaging
than indirect speech.
1.2
The vividness distinction between direct and indirect speech in
verbal communication
Indeed, linguists have observed that direct speech usually coincides with more vivid
depictions of the reported speaker‟s voice, facial expressions and gestures as compared to
indirect speech (e.g., Chafe, 1982; Clark & Gerrig, 1990; Labov, 1972; Schiffrin, 1981;
Tannen, 1982, 1986, 1989). Moreover, in English, the more frequent use of be-like, as a
quoting device for direct speech (e.g., Mary was like, „I‟m so hungry‟) but not indirect
speech, also indicates a rising communicative demand for paralinguistic demonstrations for
the former. In recent years, be-like has become more frequently used and has almost
replaced the conventional said as the alternative quotative of choice for direct speech
among younger generations (Blyth, Recktenwald, & Wang, 1990; Cukor-Avila, 2002;
Ferrara & Bell, 1995; Tagliamonte & D'Arcy, 2004; Tree & Tomlinson, 2008). Unlike
said which can only express something „in words‟, be-like can be used to describe an event,
feeling, or person using both linguistic and paralinguistic information, especially when it is
difficult to describe in words. In this sense, the increasing preference for be-like direct
Page | 13
speech may indicate a rising trend in using paralinguistic depictions (which are hard to
express in words) when reporting direct speech.
This trend is consistent with the
demonstration pragmatic function of direct speech.
In addition to the observational evidence, Clark and Gerrig‟s demonstration theory has also
been experimentally tested under laboratory environments. In Wade and Clark‟s (1993)
study, participants first watched short videotaped conversations, which they then had to
orally report to a naïve listener who was seated on the opposite side of an opaque screen.
Importantly, the narrators were either instructed to amuse their listeners (being entertaining)
or to recount the conversations as accurately as possible (being accurate). Analyses of the
participants‟ speech recordings indicated that narrators who were instructed to be
entertaining produced a greater proportion of direct-speech quotations, whereas narrators
who were instructed to be accurate produced a nearly even distribution of direct and
indirect speech quotations. In other words, language producers are more likely to employ
direct-speech quotations when communicative demands (such as the instruction to be
entertaining) encourage the use of a more vivid, demonstrative reporting style.
A
longitudinal study by Goodell and Sachs (1992) indicates that the pragmatic distinction
between direct and indirect speech is learned over many years and is mastered in adulthood.
They showed that 6-year-olds, although they may have acquired the knowledge of indirect
speech, were more likely to use direct speech than adults when retelling conversational
narratives. However, they used direct quotations one after another, in a more „self-centred‟
manner – in the sense that they did not consider which reporting style to use in order to
make listeners understand the story more effectively. Using direct speech simply relieved
them from a heavy processing load in summarising the conversationally conveyed
information (which requires using indirect speech).
Adults, on the other hand, used
indirect speech to summarise information, and saved direct speech to depict „highlights‟ of
the story. Their retelling was more „listener-centred‟ – in the sense that it attempted to
convey information effectively, taking the listener‟s vantage point into account.
In sum, both the observational and the empirical evidence support Clark and Gerrig‟s
demonstration theory of direct speech. They show that in verbal communication, direct
speech is used to provide vivid paralinguistic demonstrations of the reported speech acts
and that it is more vivid and engaging than indirect speech which provides mere
descriptions of what was said.
Page | 14
1.3
The representational consequences of the vividness distinction
in language comprehension
The vividness distinction between direct and indirect speech is obvious during verbal
communication which is rich in paralinguistic information. Our research question is: How
are direct speech and indirect speech mentally represented during language comprehension,
e.g., silent reading of text, where little paralinguistic information is available?
Classic proposals such as the „amodal symbol system‟ argue that language is represented in
arbitrary, amodal symbols that are independent from the brain‟s modal systems for
perception, action and introspection (e.g., Chomsky, 1957; Fodor, 1975; Pylyshyn, 1984).
For example, the sentence Bo opened the book can be converted into a propositional
representation (e.g., Kintsch, 1998; Kintsch & Vandijk, 1978) such as [OPENED[BO,
BOOK]]. The problem with such proposals is that those amodal, symbolic representations
are arbitrarily linked to the perceptual states that produce them. Just as the printed word
“book” has no systematic similarities to books that exist in the physical world, the amodal,
symbolic representation for book has no systematic, physical similarities to the books that
we perceive. Thus, just as the printed words “book” and “dictionary” are not necessarily
more similar than the printed words “book” and “hospital”, the amodal, symbolic
representations for book and dictionary are not necessarily more similar than the amodal
representations for book and hospital. The lack of grounding in the brain‟s modal systems
for perception, action and introspection therefore introduces the transduction problem:
How can amodal symbol systems map modality-specific perceptual states onto amodal
symbols?
By proposing an amodal, arbitrary representation of language, the amodal
symbol systems fail to provide a satisfactory account of such a transduction process.
Moreover, the lack of cognitive or neural evidence has also questioned the existence of
such a transduction process. A related shortcoming of the amodal symbol systems is the
symbol grounding problem (Harnad, 1990; Searle, 1980), i.e., how arbitrary, amodal
symbols represent the physical similarities or distinctions of referents in reality. This
problem can be illustrated by Searle‟s (1980) Chinese Room problem: Imagine you are in a
foreign country like China and unfortunately you do not speak Chinese. You wish to
understand Chinese, so you open up a dictionary, trying to study the meanings of Chinese
words. However, these words are also explained in Chinese which are just other amodal,
arbitrary symbols to you. As a result, you still do not understand Chinese. Metaphorically,
Page | 15
we can view those Chinese definitions as language representations (meanings of language),
and the Chinese words you wish to understand as language itself.
If the language
representations are arbitrary (e.g., Chinese explanations) and do not represent the physical
similarities or distinctions of their referents in the physical world, they would be
ineffective in indexing the physical referents (i.e., the meanings) of language (e.g., Chinese
words) in the physical world. In addition to the theoretically unresolved problems that
have been discussed above, the amodal symbol systems also face a shortage of empirical
support. In order to verify the existence of amodal symbol systems, we should observe
that the full spectrum of language representations are stored and processed in the same
brain area that is independent from the sensori-motor circuits (i.e., the supra-modal
language areas should exist). At the behavioural level, we should observe that language
comprehension stays independent from (i.e., has no interaction with) concurrent perception,
actions or introspection. However, over the last 20 years, there has been little direct
empirical evidence which verifies the above predictions. By contrast, researchers have
frequently discovered that language representations actually have sensori-motor
characteristics, and that language comprehension routinely interacts with concurrent
perception or actions (details discussed below).
These findings oppose the amodal,
symbolic account of language representations and are in favour of a modality-dependent,
„embodied‟ nature of language representations.
The „embodied‟ theories propose that language representations are essentially grounded in
our sensory and motor experiences of the environment, the body and the mind, and that
language comprehension is also implemented at this level via simulations (i.e., partial reenactments of perceptual, motor and introspective states acquired during experience with
the world, body and mind), making language representations well suited for interacting
with perception, actions and introspection (e.g., Barsalou, 1999, 2008; Zwaan, 2004). The
grounding of language representations in our brain‟s modal systems resolves the
transduction problem and the symbol grounding problem by making no fundamental
distinction between language representations and the corresponding perceptual states and
actions2. Under an embodied framework, we should observe that language representations
2
This is not to say that language representations and sensori-motor mental states are exactly the same in
every single aspect. The idea is that they are qualitatively equivalent in the sense that they are both modalityPage | 16
are modality-specific and that that they are stored and implemented in different brain areas
that we use for perception, action and introspection. At the behavioural level, we should
observe interactions between language comprehension and concurrent perception or action.
Indeed, behavioural and neuroimaging studies indicate that language representations are
flexible, distributed representations comprised of sensori-motor characters.
At the
behavioural level, numerous studies show that language comprehension routinely activates
sensory and motor representations, and that mental simulation of such modality-dependent
representations is an integral part of natural language comprehension (for a review, see
Barsalou, 2008). For example, researchers showed that individuals mentally simulate
objects‟ orientations (Stanfield & Zwaan, 2001) and shapes (Zwaan, Stanfield, & Yaxley,
2002) that are linguistically implied in sentences (e.g., John put the pencil in the cup
implies a vertical rather than a horizontal orientation of the pencil; The ranger saw the
eagle in the sky implies an eagle with its wings stretched out rather than folded).
Immediately after reading such sentences, subjects responded differentially to illustrations
of objects depicting congruent or incongruent orientations or shapes. These findings
suggest that language representations consist of modality-specific perceptual features (e.g.,
visual features of orientations and shapes) depending on specific sensory experiences (e.g.,
an eagle with its wings stretched out) that we acquired in the linguistically described
situations (e.g., where we see an eagle flying in the sky). In the motor domain, evidence
also indicates that understanding linguistic descriptions of actions recruits motor
representations that can be characterised in directional movements (e.g., Close the drawer
implies a movement away from the body; cf. Glenberg & Kaschak, 2002) or rotational
movements (e.g., Eric turned down the volume implies an anticlockwise rotation; cf.
Zwaan & Taylor, 2006); such motor simulations are found to interfere with individuals‟
concurrent motor responses that are required by a secondary task during reading. For
example, Glenberg & Kaschak asked subjects to make a sentence sensibility judgement
after reading sentences that imply either a towards-the-body (e.g., Open the drawer) or an
away-from-the-body (e.g., Close the drawer) directional movements. The subject had to
make responses by pressing buttons that were either near or far away from him/herself,
thereby making a towards-the-body or an away-from-the-body movement. They found
that subjects were faster in making responses when their button-pressing movements were
congruent (along the towards-vs.-away dimension) with the movements that were implied
specific. This may include the possibility that they may differ from each other to some extent on certain
quantitative dimensions (e.g., amplitude).
Page | 17
in the previously read sentences. The findings suggest that motor representations of
directional movements can be activated during language comprehension, and interfere with
language comprehenders‟ concurrent motor movements. At the neural level, researchers
attempted to map out the neural activations in relation to concepts of different modalities.
For example, neuroscientists observed that comprehension of actions words that are
semantically related to different parts of our body (e.g., „lick‟, „pick‟ and „kick‟) elicited
early (100-200 ms) neural activations in the corresponding motor and premotor cortices
(see Pulvermuller, 2005 for a review): Leg-related words activated brain areas that are
overlapping and/or adjacent to brain areas involved in foot movements and similar
relationships were also observed between comprehension of arm-related and face-related
words and neural activity in finger-movement and tongue-movement brain areas. These
observations show that language/conceptual representations are indeed stored in distinct
sensory and motor brain areas depending on specific sensory and motor experiences we
acquired during language/conceptual acquisition.
With respect to our research question, the amodal vs. embodied views of language
representations make distinct predictions regarding how direct speech and indirect speech
are mentally represented in language comprehension such as silent reading. If we adopt
the amodal view of language representations, we would predict that direct speech (e.g.,
Mary said, „I‟m hungry‟) and indirect speech (e.g., Mary said that she was hungry) are
represented in the same way.
For instance, they could be converted into identical
propositional representations such as [SAID[MARY, WAS[MARY, HUNGRY]]]. This
hypothesis seems to be philosophically implausible - if direct speech and indirect speech
are represented in exactly the same way, what is the point of having two of them in written
language, simply as the legacy from their verbal siblings? A recent study by Bohan,
Sanford, Cochrane, and Sanford (2008) demonstrates that direct speech enhances the depth
of processing in silent reading as compared with indirect speech. They employed a change
detection paradigm: They presented a sentence twice but one critical word was changed to
a semantically-similar word during the second presentation (e.g., I‟ll just have to wait
outside for my flatmate  roommate to let me in). They found that readers were more
likely to detect such subtle changes when the sentence was presented in a direct speech
quotation (e.g., She said to her boyfriend: “I‟ll just have to wait outside for my flatmate 
roommate to let me in.”) as opposed to when it was presented in an indirect speech clause
Page | 18
(e.g., She said to her boyfriend that she would just have to wait outside for her flatmate 
roommate to let her in.). Their results suggest that direct and indirect speech are processed
differently during language comprehension. Moreover, most of us tend to hear an intuitive
„inner voice‟3 particularly during silent reading of direct speech but not indirect speech.
Such differential „inner voice‟ experiences in understanding the two reporting styles also
suggest that they may be mentally represented in distinct manners, at least within the
auditory domain. In fact, this voice-related representational distinction between direct and
indirect speech in silent reading is naturally predicted by the contrasting, embodied views
of language representations. In line with these proposals, the mental representations of
direct speech and indirect speech should be grounded in the distinct sensori-motor
experiences we acquired when using or perceiving them during verbal communication. In
other words, the vividness distinction between direct and indirect speech in verbal
communication (i.e., direct speech is more vivid than indirect speech) should underlie
language comprehension of written direct and indirect speech: Comprehension of the
former should be grounded in mental simulations (re-enactments) of more vivid
demonstrations which consist of, e.g., depictions of reported speakers‟ voices; this could
explain our intuition of hearing an „inner voice‟ during silent reading of direct speech. By
contrast, comprehension of indirect speech should be grounded in mental simulations of
descriptions which are less likely to contain reported speakers‟ voices.
1.4
Thesis hypothesis and motivations
Surprisingly,
although
naturally predicted
by embodied
theories
of
language
comprehension, there has been little objectively measurable confirmation of this „inner
voice‟ experience during comprehension of direct speech to date.
Previous studies have shown that processing of sound-related concepts recruits our
auditory system (e.g., Kellenbach, Brett, & Patterson, 2001; Pecher, Zeelenberg, &
Barsalou, 2003). For instance, Pecher et al. (2003) found that verifying a property in the
3
The „inner voice‟ here refers to re-enactment of higher-level, supra-segmental acoustic features (e.g.,
Abramson, 2007; Alexander & Nygaard, 2008; Kosslyn & Matt, 1977; Kurby, Magliano, & Rapp, 2009),
rather than to lower-level phonological/prosodic encoding (e.g., Abramson & Goldinger, 1997; Ashby &
Clifton, 2005; Ashby & Rayner, 2004) which presumably occurs during silent reading of both direct and
indirect speech.
Page | 19
auditory modality (e.g., BLENDER-loud) became slower after verifying a property in a
different modality (e.g., CRANBERRIES-tart) than after verifying a property in the same
modality (e.g., LEAVES-rustling). These findings suggest that switching modalities in
conceptual processing introduces a cost, analogously to the cost of switching modalities in
perception.
Although their work did not focus on auditory simulations per se, it
demonstrates mental simulations in the auditory modality. In the same vein, Lynott and
Connell (2009) collected ratings to describe a set of 423 adjectives along the five
perceptual modalities (i.e., visual, haptic, auditory, olfactory and gustatory). Their data
indicate that sound-related object properties (e.g., beeping and meowing) are perceived
predominantly through the auditory sense.
Using simultaneous recording of
electroencelphalography (EEG) and fMRI, Kiefer, Sim, Herrnberger, Grothe, and Hoenig
(2008) demonstrated that within just 150 ms, recognition of words denoting sound-related
objects (e.g., “telephone”) activates posterior superior and middle temporal gyri
(pSTG/MTG) that are also activated during sound perception. Their results provide neural
evidence for mental simulations of sounds during conceptual processing. These studies,
although not directly related to the „inner voice‟ phenomenon, suggest the possibility that
language comprehension of speech-related information may recruit the auditory system
and result in an „inner voice‟ experience. Moreover, studies on auditory imagery in
reading (e.g., Abramson, 2007; Alexander & Nygaard, 2008; Kosslyn & Matt, 1977;
Kurby et al., 2009) demonstrate that listening to voice samples incurs an „inner voice‟
experience during reading. For example, Kosslyn and Matt (1977) demonstrated that
readers recode written material into a format analogous to that of the actual speech of the
writer after listening to an audio sample of the writer‟s voice. They showed that readers
read faster when the written material was believed to be authored by a fast-speaking person
and read more slowly when the purported author spoke slowly. These demonstrations of
„inner voice‟ experiences, although not directly related to the distinction between direct
and indirect speech, provide evidence of an „inner voice‟ in the context of language
comprehension.
Thus, this doctoral research set out to confirm the „inner voice‟ experience in
comprehension of direct speech by hypothesising that individuals are more likely to
mentally simulate the reported speakers‟ voices (among many other aspects) in
Page | 20
comprehension of direct speech as opposed to indirect speech. This research had the
following theoretical and empirical motivations.
First, this doctoral research was motivated to experimentally validate Clark and Gerrig‟s
(1990) demonstration theory from the perspective of language comprehension. Since the
formation of the theory, the supporting evidence has predominantly been provided via
observational analyses of conversations (e.g., Holt, 1996; Tree & Tomlinson, 2008) and
corpora (e.g., Tagliamonte & D'Arcy, 2004), and via experimental testing but only from
the perspective of language production (Wade & Clark, 1993). It is still unknown whether
the pragmatic distinctions between direct and indirect speech (i.e., demonstrations versus
descriptions) are also reflected in language comprehension such as silent reading of written
text. If this research empirically verified the intuitive „inner voice‟ experience in language
comprehension of direct speech versus indirect speech, it would further validate Clark and
Gerrig‟s (1990) proposal from the perspective of language comprehension, suggesting that
the direct speech reporting style is inherently more vivid than the indirect speech reporting
style regardless of language modality.
Second, this doctoral research was motivated to extend the embodied theories of language
comprehension in three respects: (1) to explore mental simulations within the auditory
perceptual modality which has received less attention in the literature; (2) to explore
mental simulations of language pragmatics, specifically in relation to semanticallyequivalent but pragmatically-distinct reporting styles; (3) to explore the underlying neural
correlates of mental simulations of voices in natural language comprehension, using stateof-the-art neuroimaging methodologies.
First of all, this research was motivated to extend the empirical demonstrations of mental
simulations to the auditory perceptual modality and to the sentence/discourse level.
During the last decade, the empirical testing of embodied theories has predominantly
focused on the visuo-spatial domain (e.g., Connell, 2007; Fincher-Kiefer, 2001; Kaup,
Ludtke, & Zwaan, 2006; Richardson & Spivey, 2000; Stanfield & Zwaan, 2001; Zwaan et
al., 2002) and on the motor domain (e.g., Glenberg & Kaschak, 2002; Kaschak & Glenberg,
2000; Kaschak et al., 2005; Richardson, Spivey, Barsalou, & McRae, 2003; Zwaan,
Madden, Yaxley, & Aveyard, 2004; Zwaan & Taylor, 2006); for a review, see Fischer &
Zwaan, 2008). There has been little evidence of auditory simulations during language
Page | 21
comprehension, especially at the sentence/discourse level (studies have only examined
mental simulations of sound-related concept words, e.g., Kellenbach, Brett, & Patterson,
2001; Kiefer, Sim, Herrnberger, Grothe, & Hoenig, 2008). Such a bias towards the visuospatial and motor domains may have been due to the urge to verify the embodied
hypothesis of language comprehension rather than to comprehensively explore such
simulation phenomena. However, in order to delineate the full spectrum of embodied
phenomena in language comprehension, it is important to extend our current scope of
empirical testing to other sensory modalities and to the sentence/discourse level which
have not been thoroughly investigated in the literature. This research, as an example of
such extensions, explored mental simulations of voice representations during natural
discourse comprehension of direct and indirect speech.
Secondly, this research was motivated to pioneer mental simulations of pragmatics during
language comprehension. The research so far has mostly focused on mental simulations in
relation to the semantics of language.
For example, Glenberg and Kaschak (2002)
examined how sentences which imply different action directions (e.g., „Close the drawer‟,
which implies action away from the body versus „Open the drawer‟, which implies action
towards the body) modulate mental simulation of motor representations during silent
reading.
They found that participants had difficulty making a sensibility judgement
requiring a body response in the mismatching direction. However, little is known about
how
semantically-equivalent
but
pragmatically-distinct
linguistic
information
is
understood during language comprehension. For example, although direct speech Mary
said: „I‟m hungry‟ and indirect speech Mary said that she was hungry convey virtually the
same semantic meaning, the former is likely to coincide with vivid paralinguistic
demonstrations of the reported speech act whereas the latter is reported in a less vivid
fashion. Thus, by demonstrating how the vividness distinction between direct and indirect
speech underlies language comprehension, this doctoral work would open up a new
research dimension of the embodiment of language pragmatics.
Finally, given the wide range of neuroimaging facilities that are available within the Centre
for Cognitive Neuroimaging at the University of Glasgow, this research was motivated to
apply the state-of-the-art neuroimaging techniques in psycholinguistic research. Previous
research on mental simulations in language comprehension has mostly employed
behavioural paradigms. For example, some employ reaction time-based paradigms which
Page | 22
consist of language comprehension and a secondary recognition task.
Using such
paradigms, researchers examine how language comprehension affects individuals‟
response latencies to the relevant probe stimuli (e.g., a probe picture) that were made
relevant by the previous text. Although those behavioural methodologies have their merits,
they are limited in specifying the underlying neural correlates of mental simulations during
natural language comprehension (without a secondary task). In order to fill this gap, this
doctoral research combined event-related functional magnetic resonance imaging (fMRI)
with eye-tracking to investigate the neural correlates of the hypothesised „inner voice‟
experience during comprehension of direct versus indirect speech. This interdisciplinary
approach could inspire new applications in psycholinguistic research, advancing our
understanding of the underlying architectures and neural circuitry of language
comprehension.
1.5
Thesis overview
The following is a condensed overview of the empirical work that is presented in this
thesis.
The more in-depth discussions regarding the theoretical backgrounds and the
interpretations of the results will be presented in the individual chapters. Overall, this
thesis is presented in three parts. The first part reports functional magnetic resonance
imaging (fMRI) evidence of an „inner voice‟ experience in comprehension of direct versus
indirect speech.
The second part reports behavioural evidence of this „inner voice‟
experience, aiming to characterise the underlying mental representations into acoustic
parameters of pitch, loudness and rate. The third part of the thesis ventured beyond the
„inner voice‟ phenomena and explored mental simulations of facial expressions during
silent reading of direct versus indirect speech.
Page | 23
Part
1:
Differential
brain
activity
in
voice-selective
areas
during
comprehension of direct versus indirect speech
Chapter 2 tested our hypothesis by measuring the „top-down‟ activations of the auditory
cortex4 during silent reading of direct as opposed to indirect speech. We reason that if
individuals are more likely to simulate „inner voice‟ representations during silent reading
of direct speech rather than indirect speech, the brain areas which are specialised in
processing human voices should become more active when reading the former as opposed
to the latter. In other words, we should observe higher brain activity within the human
voice-selective areas (Belin, Zatorre, Lafaille, Ahad, & Pike, 2000) 5 during silent reading
of direct as opposed to indirect speech. Combining functional magnetic resonance imaging
(fMRI) with eye-tracking in human volunteers, we show that silent reading of direct as
opposed to indirect speech engenders significantly higher brain activation in the voiceselective areas of the right auditory cortex. The findings suggest that readers are indeed
more likely to engage in mental simulations (or spontaneous imagery) of reported speakers‟
voices when reading direct speech as opposed to meaning-equivalent indirect speech, as
part of a more vivid representation of the former. The results are discussed in comparison
with research on auditory imagery and lip-reading, and were interpreted in line with the
demonstration theory of direct speech (Clark & Gerrig, 1990) and the embodied hypothesis
of language comprehension (Barsalou, 1999).
Chapter 3 explored the supra-segmental nature of the observed „inner voice‟ experience in
comprehension of direct rather than indirect speech. We measured the „top-down‟ brain
activity in the voice-selective areas when listening to monotonous direct speech utterances
versus monotonous indirect speech utterances. We reason that if direct speech is grounded
in mental simulations of supra-segmental acoustic information, listeners would have to
mentally simulate such information when listening to monotonous direct speech in which
such information is minimised; in contrast, listeners would be less likely to simulate such
supra-segmental acoustic information when listening to monotonous indirect speech as the
4
„Top-down‟ means that the activation of the auditory cortex is NOT triggered by „bottom-up‟ auditory
stimulation itself. In our case, it is triggered by higher-level language comprehension.
5
The human voice-selective brain areas include the superior temporal gyrus/superior temporal sulcus
(STG/STS) areas bilaterally. The maximum of sensitivity to vocal sounds was located along the upper bank
of the central part of the right STS area.
Page | 24
latter is not grounded in vivid vocal depictions. Thus, we should observe higher „top-down‟
brain activity in the voice-selective areas when listening to monotonous direct speech
utterances as compared to monotonous indirect speech utterances. Using fMRI, we found
that listening to monotonous direct speech utterances indeed elicited significantly higher
brain activations in the voice-selective areas of the right auditory cortex than listening to
monotonous indirect speech utterances.
Follow-up rating studies on contextual
congruency revealed that monotonous direct speech utterances were perceived as
incongruent with their preceding contexts whereas monotonous indirect speech utterances
were perceived as contextually congruent. Parametric modulations showed that part of the
observed differential brain activations were significantly accounted for by such contextual
congruency, but not by the stimuli‟s acoustical characteristics or their vividness perceived
by listeners. The results indicate that within a given context, listeners routinely expect
vivid vocal depictions for direct speech but not for indirect speech; they are more likely to
mentally simulate such vivid vocal depictions (i.e., enriched supra-segmental acoustic
representations), when listening to monotonous direct speech (contextually incongruent) as
opposed to monotonous indirect speech (contextually congruent). The observed brain
activations were reconciled with those observed in silent reading (Chapter 2) since they
were located in virtually the same brain areas across the two studies. Such a betweenstudy consistency suggests that the „inner voice‟ experience we observed in silent reading
may have also been supra-segmental in nature (e.g., intonation and emotional prosody).
Moreover, the current study also provides novel demonstrations of top-down processing
during intelligible speech perception which is modulated by pragmatically distinct
linguistic information.
Part 2: The acoustic aspects of the ‘inner voice’ in comprehension of direct
versus indirect speech
Chapters 2 and 3 together provide brain-activation evidence that individuals are indeed
more likely to mentally simulate enriched supra-segmental acoustic representations of the
reported speakers‟ voices during language comprehension of direct speech as opposed to
indirect speech. Chapters 4 – 6 attempted to further decompose such representations in
the acoustic parameters of pitch, loudness and rate, using behavioural methodologies.
Page | 25
Chapter 4 investigated whether the „inner voice‟ in reading direct speech contains pitch
representations of the reported speakers‟ voices. We reasoned that if the „inner voice‟
experience in comprehension of direct rather than indirect speech entails pitch information,
it may affect the comprehenders‟ articulation (in modulations of pitch) or voice perception
(facilitation on pitch-congruent voice stimuli). First, using an oral-reading paradigm, we
observed that oral readers spontaneously adjusted their own pitch in accordance with the
reported speaker‟s emotional arousal and gender that were implied in the previously read
linguistic context, however only during oral reading of direct speech but not indirect
speech. Moreover, we found that the differential pitch modulations during oral reading of
direct versus indirect speech were masked when linguistic focus was exerted upon reported
speakers‟ voices; such a „voice-selective‟ masking effect indicates the underlying
representations of such pitch modulations are likely to be voice-related in nature. The
findings suggest that oral readers are more likely to mentally simulate reported speakers‟
pitch representations during oral reading of direct as opposed to indirect speech. Next, we
developed three reaction time-based paradigms to examine mental simulations of pitch
during silent reading of direct versus indirect speech. We examined the effects of reported
speakers‟ genders on the participants‟ recognition latencies to spoken words that were
recorded from either a same-gender (as the reported speaker) speaker or a different-gender
speaker (the stimuli only significantly differed in pitch). Using a serial visual presentation
paradigm in which the written critical words (for silent reading) and the spoken probe
words (for recognition) were presented simultaneously, we found that the participants
responded more slowly to the different-gender spoken words than to the same-gender
spoken words, however only during silent reading of direct speech quotations that were
reported from female speakers (e.g., She said: „…‟). The results show that silent readers
are more likely to mentally simulate the gender-accordant pitch representations during
silent reading of direct speech rather than indirect speech, but only when the reported
speakers are female; this discrepancy between male and female reported speakers may be
interpreted in the acquired experience that female speakers are generally more emotional
and expressive when reporting direct speech during verbal communication (e.g., Kring &
Gordon, 1998; Robinson & Johnson, 1997). Although interesting, this interpretation is
pending further confirmation with fully-counterbalanced materials and more carefullymanipulated auditory stimuli.
Page | 26
Chapter 5 examined whether the „inner voice‟ experience entails reported speakers‟
loudness information during comprehension of direct versus indirect speech. Similar to
our previous investigations on pitch simulations, we examined the effects of the purported
loudness simulations on participants‟ articulation and voice perception.
In two
experiments, we asked participants to orally (Experiment 7) or silently (Experiment 8,
reaction time-based method) read written stories that contained either direct speech or
indirect speech quotations. The context preceding those quotations described a situation
which implied either a loud-speaking or a quiet-speaking quoted protagonist. We found
that this context manipulation modulated the participants‟ reading loudness during oral
reading of direct speech but not indirect speech. The findings suggest that oral readers
may mentally simulate loudness representations during oral reading of direct rather than
indirect speech. However, in silent reading, the results were not very straightforward. We
examined the effects of the contextually-described loudness of the reported speakers on the
participants‟ recognition latencies to spoken words that were either loud or quiet in volume.
We found that the loudness (volume) levels of the spoken words did not significantly affect
the participants‟ response times, suggesting that loudness representations may not be
encoded in vocal volume during silent reading of direct or indirect speech. Unexpectedly,
we found a significant context effect in the indirect speech condition rather than the direct
speech condition: the participants responded to the spoken words significantly faster when
reading a loud-speaking context as opposed to a quiet-speaking context, but only during
silent reading of indirect speech rather than direct speech. This context effect may reflect
modulations of response times by the differential emotional intensities that are confounded
with reported speakers‟ loudness; such RT modulations by emotional intensity may be
cancelled out during silent reading of direct speech due to the integration of the simulated
voice representations with the perceived voices of the spoken words. Reconciled the
current findings with the literature, it was speculated that speakers‟ loudness may not be
encoded in vocal volumes, but perhaps in other forms.
Chapter 6 studied whether the „inner voice‟ experience represents reported speakers‟
speaking rates during comprehension of direct versus indirect speech. In two experiments,
we asked participants to either orally (Experiment 9) or silently (Experiment 10, eyetracking) read written stories that contained either a direct speech or an indirect speech
quotation. The context preceding those quotations described a situation that implied either
Page | 27
a fast-speaking or a slow-speaking quoted protagonist.
We found that this context
manipulation affected reading rates (in both oral and silent reading) for direct speech
quotations, but not for indirect speech quotations. The findings suggest that readers are
more likely to engage in mental simulations of the reported speakers‟ voices when reading
direct as opposed to indirect speech, and that the speaking rate is an important acoustic
aspect of such voice representations.
Part 3: The visual aspect of the direct-speech representations: Mental
simulations of facial expressions?
We have obtained neuroimaging evidence which suggests that individuals are indeed more
likely to mentally simulate enriched supra-segmental acoustic representations of reported
speakers‟ voices during comprehension of direct speech as opposed to indirect speech. We
have also obtained behavioural evidence which indicates that the simulated voice
representations in reading of direct versus indirect speech can be characterised in
modulations of acoustic parameters such as speaking rate.
However, in verbal communication, reported speakers‟ facial expressions and gestures are
usually co-demonstrated with voices during direct speech reporting. To comprehensively
examine the grounding of language in such multi-dimensional demonstrations of direct
speech, Chapter 7 ventured beyond the investigations of „inner voice‟ experience and
explored whether reported speakers‟ facial expressions are mentally simulated during
comprehension of direct speech rather than indirect speech. We examined the effects of
direct versus indirect speech quotations on the recognition latencies to graphic emoticons
which were made relevant by the contexts preceding those quotations. Participants silently
read written stories containing either a direct speech or an indirect speech quotation, each
followed by a graphic emoticon, and indicate whether the picture was congruent with the
story they had just read. The facial expression illustrated in the emoticon either matched or
mismatched the expression of the reported speaker implied in the story. Three expressions
(happy, surprised and angry) were examined. Because participants frequently confused
the surprised expression with either the happy or angry expression, we excluded the
„surprised‟-related items to ensure the validity of the to-be-examined expression-match
effects. On the remaining „surprised‟-free trials, we found that responses to the graphic
emoticons were facilitated when the illustrated facial expression matched the implied
Page | 28
expression of the reported speaker, but only after silent reading of direct speech. This
expression-match effect was „masked‟ when the linguistic focus was manipulated on
reported speakers‟ facial expressions (i.e., both direct and indirect speech quotations were
preceded with explicit descriptions of facial expressions, e.g., She smiled). Such a „faceselective‟ masking effect suggests that the representational differences underlying this
expression-match effect may be facial expression-like in nature. Despite the fact that many
contrasts did not reach significance due to the exclusion of a considerable number of
„surprised‟ trials, the result patterns were highly consistent with our hypothesis.
In
conjunction with the face-related brain activations that were additionally observed in silent
reading of direct rather than indirect speech (Chapter 2), the findings imply that silent
readers might indeed simulate reported speakers‟ facial expressions during silent reading
of direct speech more so than during silent reading of indirect speech. Nevertheless, future
research is needed to replicate the current results with a larger number of items, and using
different, unambiguous expressions.
In sum, my PhD work provides strong support for the intuitive „inner voice‟ experience we
have when reading direct speech but not indirect speech. It demonstrates that individuals
are more likely to mentally simulate enriched supra-segmental acoustic representations of
reported speakers‟ voices during comprehension of direct speech as opposed to indirect
speech. Simulated voice representations seemed to be sensitive to modulations of acoustic
parameters such as speaking rate but not pitch or loudness.
Further investigations,
however, may yet reveal differences by modulations of pitch or loudness using more
sensitive experimental paradigms.
Moreover, preliminary results suggest that facial
expressions may also be part of a more vivid, multi-dimensional representation of direct
speech in language comprehension.
The findings support Clark and Gerrig‟s (1990)
demonstration theory of direct speech from the persepective of language comprehension
and extend the empirical work on embodied theories of language comprehension (e.g.,
Barsalou, 1999) to the auditory perceptual modality and the embodiment of language
pragmatics, with the application of interdisciplinary methodologies.
Page | 29
PART 1: Differential Brain Activity in Voice-Selective Areas
during Comprehension of Direct versus Indirect Speech
Page | 30
Chapter 2 Brain ‘hears’ voices when reading direct quotations:
Silent reading of direct versus indirect speech activates the
human voice-selective areas (Experiment 1)
2.1
Introduction
As discussed in Chapter 1 (see 1.2), direct speech (as in Mary said: „I‟m hungry‟) is
assumed to entail vivid demonstrations of the reported speech acts, whereas its counterpart
indirect speech (as in Mary said that she was hungry) provides mere descriptions of what
was said (Clark & Gerrig, 1990). In verbal communication, direct speech demonstrations
are usually more vivid and perceptually engaging than indirect speech descriptions. This
vividness distinction has been observed and discussed by linguists (e.g., Tannen, 1986,
1989).
It has also been shown, for instance, that in reporting previously overheard
dialogues, speakers are more likely to employ direct rather than indirect speech when
instructed to be entertaining to a listener (Wade & Clark, 1993). In their study, participants
first watched short videotaped conversations, which they then had to orally report to a
naïve listener who was seated on the opposite side of an opaque screen. Importantly, the
narrators were either instructed to amuse their listeners (being entertaining) or to recount
the conversations as accurately as possible (being accurate). Analyses of the participants‟
speech recordings indicated that narrators who were instructed to be entertaining produced
a greater proportion of direct speech quotations, whereas narrators who were instructed to
be accurate produced a nearly even distribution of direct and indirect speech quotations. In
other words, language producers are more likely to employ direct speech quotations when
communicative demands (such as the instruction to be entertaining) encourage the use of a
more vivid, demonstrative reporting style.
However, little research so far has addressed the question of how the two reporting styles
are represented in language comprehension, particularly during silent reading of text where
no auditory stimulation, or visual stimulation other than text, is present. While many of us
share the intuition of hearing an „inner voice‟ during silent reading of direct rather than
indirect speech, there has been hardly any objectively measurable confirmation of this
experience so far. This is surprising, especially given that recent embodied theories (e.g.,
Barsalou, 1999, 2008) propose that language comprehenders mentally simulate
linguistically described situations based on generalised perceptual experiences they have
Page | 31
acquired in the past. Such theories suggest that, even during silent reading of text, direct
speech may be more likely to activate „audible-speech‟-like representations than indirect
speech. In other words, readers may be more likely to mentally simulate the reported
speaker‟s voice (or aspects thereof) during silent reading of direct as opposed to indirect
speech.
One way to test this hypothesis is using functional magnetic resonance imaging (fMRI) to
measure „top-down‟ activation of the auditory cortex during silent reading of direct versus
indirect speech. The fMRI technique has been developed based on the idea that local
cerebral blood flow and blood oxygenation reflects local neural activity (Roy &
Sherrington, 1890): It measures changes in local deoxyhemoglobin concentration, Bloodoxygen-level-dependence (BOLD), which is reflected in alterations in signal intensity of
magnetic resonance images (Ogawa & Lee, 1990; Ogawa, Lee, Kay, & Tank, 1990;
Ogawa, Lee, Nayak, & Glynn, 1990). Brain images characterising changes of BOLD
signals thereby indicate modulations of neural activity within certain brain areas as a
function of manipulated psychological processes. From the fMRI literature, it is known
that certain brain areas in the auditory cortex are selectively sensitive to human voices
when stimulated „bottom-up‟ via auditory sound clips (Belin et al., 2000). Belin and
colleagues found that bilateral brain areas along the upper bank of the superior temporal
sulcus (STS) showed greater brain activity when participants listened passively to vocal
sounds (e.g., coughing, laughter and speech), than to non-vocal environmental sounds (e.g.,
telephone ringing). Their findings provide us with clearly defined regions of interest for
our investigations of the activation of voice-related mental representations during on-line
silent reading. Moreover, studies on non-verbal (Bunzeck, Wuestenberg, Lutz, Heinze, &
Jancke, 2005; Yoo, Lee, & Choi, 2001) and verbal (Jancke & Shah, 2004) auditory
imagery, as well as experiments on visual speech perception (also known as lip-reading)
(Calvert et al., 1997; MacSweeney et al., 2000), indicate that areas within the auditory
cortex are also prone to „top-down‟ activation without external auditory stimulation.
Furthermore, Spitsyna and colleagues (2006) showed that neural pathways for silent
reading and speech perception converge in the superior temporal sulcus (STS) regions of
the auditory cortex, suggesting the possibility that silent reading might recruit some
auditory processing. Hence, if readers are more likely to engage in mental simulations of
reported speakers‟ voices during silent reading of direct speech, voice-selective areas in the
Page | 32
auditory cortex (Belin et al., 2000) should display enhanced „top-down‟ activation as soon
as readers come across a direct speech quotation (as opposed to a meaning-equivalent
indirect speech statement) in written text. The following event-related fMRI experiment
aimed at testing this hypothesis. Participants silently read a number of short written stories
for comprehension while their eye-movements and brain activations were simultaneously
monitored.
2.2
Method
2.2.1 Participants
In total, 26 adult participants were recruited and scanned. All of them were native English
speakers with normal vision and hearing and having no learning or reading disabilities nor
history of neurological or psychiatric disorders. Ten participants had to be excluded from
analysis due to either (a) no clear response in the voice localiser task (see below) and/or
excessive head-movements during scanning (8 subjects), (b) eye-tracking data loss (1
subject), or (c) less than 70% answering accuracy on comprehension questions (1 subject).
Data from the remaining sixteen participants (age 18 – 44 years, six males and ten females)
were used for the final analyses. All of them were right-handed except for one female
subject.
2.2.2 Stimuli and design
Silent reading stimuli
Ninety short stories with different protagonists (indicated by different names) were
prepared as reading materials (Appendix 1). Each story started with two declarative
sentences to set up a scenario (e.g., PhD student Ella was summoned to her supervisor
Jim‟s office to give a report on her current progress. Ella asked for an extension but Jim
looked concerned.). This was followed by either a direct or an indirect speech quotation
(e.g., He said: „Hmm, we really need those data in by next month for that conference.‟ or
He said that they really needed those data in by next month for that conference.). The
reported quotations in both conditions (underscored in the above examples) were
equivalent in terms of linguistic content. Comprehension questions (e.g., Was Ella Jim‟s
Page | 33
PhD student?) were also prepared for 23 stories (ca. 25%) to assess participants‟ overall
comprehension accuracy and to ensure that they read the stories for comprehension.
Two lists of stimuli with counterbalanced item-condition combinations (45 direct and 45
indirect speech trials per list) were constructed. Each item appeared only once per list, but
in a different condition across lists. Half of the participants saw presentation list one and
the other half presentation list two. The presentation order of the items within each list
was randomised for each participant.
Voice localiser stimuli
For the voice localiser session (see 2.2.3 Procedure), we presented blocks of vocal sounds
and
non-vocal
sounds
provided
by
the
Voice
Neurocognition
Laboratory
(http://vnl.psy.gla.ac.uk), CCNi, University of Glasgow. These stimuli were the same as
those employed in Belin et al. (2000), and comprised both speech (e.g., spoken vowels)
and non-speech (e.g., laughing and coughing) vocal sound clips, as well as non-vocal
sound clips (e.g., telephone ringing, dog barking and bell chiming). The contrast in brain
activity elicited by vocal versus non-vocal sounds reliably localises voice-selective areas of
the auditory cortex.
2.2.3 Procedure
Participants were positioned in the scanner, wearing goggles (NordicNeuroLab, Bergen,
Norway) for visual presentation and eye-tracking (Viewpoint Eye-Tracker, Arrington
Research, Inc.) as well as MRI-compatible, electrostatic headphones (NordicNeuroLab) for
noise attenuation during fMRI scanning and for auditory presentation during the voice
localiser session (see below). After a brief eye-tracker calibration procedure, the main
reading session followed during which the participant‟s brain was scanned and their eye
movements recorded. Participants were instructed to read the text silently and carefully so
as to be able to answer comprehension questions which would follow after 25% of the
short stories they had read. The stimulus materials were presented using E-prime 2.0
(Psychology Software Tools, Inc., USA). The text was presented in a black 15-point
Courier New font on a light grey background. The whole text was centred and wrapped
within 75% width of the screen, together making the reading stimuli appear as natural as
Page | 34
possible. Each trial began with the presentation of a fixation cross in the middle of the
screen for 2 seconds, followed by the presentation of the text for reading. Each story was
presented in a sentence-by-sentence fashion.
The presentation duration per sentence
display was determined as W × 100 ms + S × 50 ms (where W refers to the number of
words and S to the number of syllables per sentence), allowing sufficient time for reading.
Mean presentation durations for the final (critical) sentence display were 7514 ms (SD =
1412 ms) and 7526 ms (SD = 1354 ms) for the direct and indirect speech condition,
respectively. About 25% of the text presentations were followed by a comprehension
question regarding the content of the preceding story. Each such question appeared in the
middle of the screen, prompting a „yes‟ or „no‟ response which participants could provide
by pressing buttons on a response box with their index or middle fingers, respectively.
Figure 1A provides a schematic illustration of the experimental trial sequence. The 90
reading trials were evenly interspersed with five „baseline‟ trials during which a plain
fixation cross was presented in the centre of the screen for 30 seconds.
After the main reading session, an anatomical scan of the participant‟s brain was
performed, followed by a brief, 10-min voice localiser scanning session. During the latter,
participants were instructed to close their eyes while listening to 20 8-second blocks of
vocal and 20 8-second blocks of non-vocal auditory stimuli presented in an efficiency
optimised, pseudo random order along with 20 blocks without stimulation acting as a
baseline (Figure 1B; cf. Belin et al., 2000).
Page | 35
Figure 1 Experimental procedures for (see Experimental Procedures), A, Silent reading task, B, Voice
localiser task.
2.2.4 MRI acquisition
Scanning was performed on a 3-T Siemens Tim Trio MRI scanner using a 12-channel head
coil (Erlangen, Germany).
Functional scans (for both the reading session and voice
localiser session) were acquired using a T2*-weighted Echo Planar Imaging (EPI)
sequence (32 slices acquired in orientation of the Sylvian fissure; TR = 2s; TE = 30 ms;
matrix size: 70 × 70; voxel size: 3 × 3 × 3 mm; FOV = 210). T1 whole-brain anatomical
scans were obtained using 3D T1-weighted magnetization prepared rapid acquisition
Page | 36
gradient echo (MP-RAGE) sequence (192 axial slices; matrix size: 256 × 256; voxel size:
1 × 1 × 1 mm; FOV = 256). The average scanning time for the whole experiment was
around 53 min per participant.
2.2.5 Data analysis
All MRI data were analysed using SPM8 (www.fil.ion.ucl.ac.uk/spm/, University College
London). Pre-processing of functional scans included (a) head motion corrections (trilinear interpolation) whereby scans were realigned to the first volume; (b) co-registration
of functional scans to their corresponding individual anatomical scans; (c) segmentation of
the co-registered scans; (d) normalisation of functional (3 mm isotropic voxels) and
anatomical (1 mm isotropic voxels) data to the Montreal Neurological Institute (MNI)
space; (e) smoothing of normalised data (8-mm Gaussian kernel).
fMRI data from the anatomical and the voice localiser scanning sessions were used to
determine the voice-selective areas in the auditory cortex of each participant at p < 0.001
(Family-Wise Error (FWE)-corrected). The group voice localiser was obtained at p <
0.001 (uncorrected, to increase sensitivity against the background of individual
differences).
For the reading scanning session, the temporal onset of a critical fMRI-event was defined
(via eye-tracking) as the temporal onset of the first fixation in the first continuous reading
of the direct or indirect speech quotation in the text; its offset was defined as the temporal
offset of the last fixation in the first continuous reading of the direct or indirect speech
quotation. Average critical fMRI-event durations amounted to 3118 ms (SD = 1175 ms)
and 3266 ms (SD = 1209 ms), for the direct and indirect speech conditions, respectively.
The 148 ms difference in the mean durations is most likely due to the fact that the indirect
speech quotations were on average 0.8 words longer than the direct speech quotations. In
fact, on a reading time per word measure, there was no appreciable difference between the
two conditions (direct speech: 204 ms per word; indirect speech: 203 ms per word; p > 0.5
by paired-samples t-test).
Hence, when minor differences in numbers of words are
controlled for, it appears that direct and indirect speech quotations were virtually identical
in terms of processing difficulty.
Page | 37
Non-critical readings (of background sentences, comprehension questions, instructions,
etc.) and events (button pressing) were specified as corresponding events in the design
matrix. The rest consisted of all fixation-cross events (including five 30-second baseline
trials and all 2-second pre-trial fixation crosses) and were regarded as baseline. The fMRI
data were mapped to the human Colin atlas (sumsdb.wustl.edu/sums/) surface in CARET
(brainvis.wustl.edu/wiki/index.php/Caret:About; (D. C. Van Essen, 2002; D.C. Van Essen,
Harwell, Hanlon, & Dickson, 2005).
The mean beta estimates within ROIs were
calculated using SPM toolbox easyROI (www.sbirc.ed.ac.uk/cyril/cp_download.html), and
submitted to 2-tailed paired-samples t-tests.
2.3
Results
The answering accuracy on the post-trial comprehension questions amounted to 83%
(direct speech condition) versus 82% (indirect speech condition). The 1% difference was
not significant (p > 0.4 by logit binomial GEE; Hardin & Hilbe, 2003).
During critical events (determined via eye-tracking, see 2.2.5 Data analysis), it was found
that the direct speech condition was associated with a greater blood-oxygen-level
dependent (BOLD) signal in voice-selective brain areas of the right auditory cortex than
the indirect speech condition. Although both conditions were more active against the
baseline, reading of direct speech elicited significantly greater activation in these areas
than reading of indirect speech (see Figure 2C). For individual Regions of Interest (ROIs),
the mean between-condition difference amounted to 0.240 ± 0.062 (SE) (2-tailed pairedsample t(15) = 3.85, p = 0.002); for the group ROI, the between-condition difference was
0.147 ± 0.045 (t(15) = 3.25, p = 0.005). Three main clusters of enhanced activity for direct
as opposed to indirect speech were located in voice-selective areas along the posterior,
middle and anterior parts of the right superior temporal sulcus (STS) (Figure 2B). In
addition, activation in brain areas other than the auditory cortex was found distributed in
the occipital lobes, superior parietal lobules and precuneus (Figure 2A). Although not
central to our hypothesis, one might speculate that activation of those areas is part of an
enriched multi-sensory mental simulation process for direct speech that also encompasses,
for instance, visual aspects of the described situation. As a whole, no region showed an
opposite pattern of activity, i.e., direct speech was always associated with a greater BOLD
signal than indirect speech. These results support the hypothesis that readers are more
Page | 38
likely to engage in vivid mental simulations of the reported speaker‟s voice (or aspects
thereof) during silent reading of direct speech as opposed to meaning-equivalent indirect
speech.
Figure 2 Illustrations of between-condition differences, A, Regions that selectively responded to silent
reading of direct as opposed to indirect speech in the whole-brain analysis, B, Regions that selectively
responded to silent reading of direct as opposed to indirect speech within the sample voice-sensitive areas
(green, n = 16), C, Mean signal change (against baseline) between conditions in the right voice area Regions
of Interest (ROIs), determined individually (left) or using the sample average (right). The single error bar in
each panel refers to the 95% CI for the between-condition difference.
2.4
Supplemental analyses
One might point out that our direct versus indirect speech materials sometimes differed in
the use of exclamations (e.g., She said: „My life has been amazing!...‟), such that 39 of our
90 direct speech items, but none of our 90 indirect speech items, employed emphasisenhancing exclamation marks. Moreover, the direct and indirect speech conditions were
Page | 39
not always parallel in terms of syntactic coordination (e.g., main clause followed by
another main clause) versus subordination (e.g., main clause followed by a subordinate
clause). Upon inspection of our stimuli, we found that the latter strictly applies to 11 items.
Of the remaining 79 (syntactically more or less parallel) items, 43 contained direct and
indirect speech sentences that can be regarded as fully equivalent in terms of syntactic
structure.
We addressed these potential concerns as follows. There were 34 items in our set of
stimuli (Appendix 1) that neither contained exclamations in the direct speech condition,
nor implied any substantial syntactic differences between direct and indirect speech (i.e.,
from the 43 items where direct and indirect speech shared the same syntactic structure, we
excluded 9 items that contained exclamations). The item numbers of these 34 items can be
found below.
We re-analysed our data (using the same threshold criteria as before)
considering only this restricted set of items, which naturally implied a considerable
reduction in statistical power due to fewer data points per participant (one third of all the
data available). Nevertheless, we found that the previously established effect (over the
whole set of items) did not completely disappear when only those 34 items were
considered: direct speech was still associated with more activation in the regions of interest
than indirect speech.
For individual Regions of Interest (ROIs), the mean between-
condition difference amounted to 0.162 ± 0.077 (SE) (2-tailed paired-sample t(15) = 2.093,
p = 0.054); for the group ROI, the between-condition difference was 0.087 ± 0.073 (the
latter was not significant, but still pointing in the right direction). We then ran analyses
comparing these 34 „uncontroversial‟ items with the remaining 56 „controversial‟ ones,
constituting the factor item group in a 2 [Item group: controversial versus uncontroversial]
× 2 [Reporting style: direct versus indirect] within-subjects ANOVA. Consistent with our
original t-test analyses over the whole set of items, we found a significant main effect of
reporting style (individual ROIs: F(1,15) = 11.602; p < .005; group ROI: F(1,15) = 6.983;
p < .02). However, the main effect of item group was far from significant (F-values < 1),
and so was the item group by reporting style interaction (F-values < 1.3) – the latter would
be expected to approach significance if exclamations and/or syntactic complexity
differences within the 56 controversial items were mainly responsible for the reporting
style effect over the whole set of items. Taken together, these analyses suggest that our
main results cannot be reduced to differences in emphasis and/or syntactic structure
Page | 40
between the direct and indirect speech conditions. Finally, the fact that we did not find any
substantial differences in reading time per word or comprehension accuracy between the
two conditions (see 2.2.5 Data analysis & 2.3 Results) may also be taken as an argument
against syntactic complexity as a potential confound – at the very least, it suggests that
potential syntactic complexity differences between direct and indirect speech were not
processing-relevant.
*The 34 items chosen for re-analysis were (item numbers in Appendix 1): 2, 3, 5, 7, 8, 10,
11, 16, 22, 25, 27, 28, 29, 30, 31, 32, 36, 38, 39, 40, 41, 48, 52, 53, 54, 55, 60, 66, 69, 70,
72, 78, 80, 88.
However, even in the subset of 34 items, many direct speech utterances used interjections
such as “Oh!”, “Hmm…” and “Ah!” which did not appear in the corresponding indirect
speech utterances. One concern is that most of the interjections, like exclamation marks,
tend to be emotionally arousing and therefore may have confounded the higher brain
activations in the right STS area that we observed in the direct speech condition.
Admittedly, we cannot rule out that direct speech tends to be associated with increased
emotionality. In fact, the reason why direct speech is more vivid or expressive than
indirect speech is that when using direct speech, reporters demonstrate or depict the
reported speech act, which may also include a demonstration of the quoted speaker‟s
emotional state. We assume that emotionality and voice cannot be ultimately disentangled
in speech because voice (just like facial expression) is an important “vehicle” of the
reported speaker‟s emotion, and emotion (or supra-segmental acoustic information that
encodes emotion in a voice) may be an important aspect of what is mentally simulated in
our experiment. For this reason, it would be implausible to claim that mental simulations
of voice during silent reading of direct speech are emotion-free. That said, it is also very
unlikely that our data reflect emotionality „per se‟.
First, a recent finding from the
literature suggests that modality-unspecific emotions are represented in the medial
prefrontal cortex (MPFC) and left pSTS (Peelen, Atkinson, & Vuilleumier, 2010) rather
than the right STS. Second, our analyses explicitly targeted regions in the brain that are
sensitive to voice, using a voice-localiser task that identified regions of interest via bottomup stimulation. Therefore we are not convinced that increased emotionality of direct
speech “on its own” would provide a feasible alternative interpretation of our data – at the
very least, voice must be taken into consideration as well. Finally, we are leaning towards
Page | 41
discussing emotional prosody (Wildgruber, Ackermann, Kreifelts, & Ethofer, 2006) as a
potential candidate for the kinds of voice-related representations that may be activated
during reading of direct as opposed to indirect speech (see Discussion below). This
interpretation thereby adequately integrates enhanced voice simulations with increased
emotionality in the direct speech condition.
2.5
Discussion
Overall, our results lent objective empirical support to the intuitive experience of an „inner
voice‟ during silent reading of written text, particularly during silent reading of direct
speech. Specifically, our experiment showed that voice-selective areas in the auditory
cortex become more active during silent reading of direct speech as opposed to meaningequivalent indirect speech. This finding could not plausibly be attributed to differences in
processing difficulty or comprehension performance because there was no significant
difference between the two conditions in terms of reading time per word or questionanswering accuracy. Other factors, the use of direct speech exclamations and potential
syntactic complexity differences in some of our direct versus indirect speech item pairs,
also failed to conclusively account for the observed brain activations.
Previous behavioural studies on auditory imagery (Abramson, 2007; Alexander & Nygaard,
2008; Kurby, Magliano, & Rapp, 2009) also observed the sort of „inner voice‟ during silent
reading. However, these studies predominantly used such an experimental setup in which
participants were presented with particular voices before the actual reading trials, followed
by explicit imagery instructions combined with cues to the identity of the to-be-imagined
speaker within the actual reading materials. This type of experimental manipulation (preexposure to specific voices and/or explicit imagery cues and instructions) arguably
encourages participants to imagine speaker-specific voices during silent reading. Hence it
is limited in determining whether auditory imagery would also occur as a result of „normal‟
language comprehension.
The current experiment focused on (1) whether silent readers activate voice-related
representations even without being encouraged to do so (reading for comprehension only)
and (2) whether these representations are modulated by different linguistic reporting styles
(direct versus indirect speech). A setup was used whereby no particular speaker was
Page | 42
introduced to participants before reading the text passages and in which participants were
in no way instructed to imagine voices (all they were asked to do was to read short stories
and to answer questions about those stories).
Each story contained a unique set of
fictitious names and characters such that participants were not (or at least not obviously)
led towards imagining any concrete, familiar voices. Combined with the fact that postreading comprehension questions focused on semantic content (distracting participants‟
attention away from the direct versus indirect speech manipulation), it seems unlikely that
our participants felt encouraged to imagine specific voices during reading. Nonetheless,
they still appeared to automatically activate voice-related perceptual representations,
particularly in response to reading direct speech quotations.
The clearly right-lateralised brain activation pattern observed in our study also suggests
spontaneous (in contrast to encouraged) activations of voice-related representations during
silent reading of direct as opposed to indirect speech.
Previous studies on auditory
imagery and on lip-reading (Bunzeck et al., 2005; Calvert et al., 1997; Jancke & Shah,
2004; MacSweeney et al., 2000; Shergill et al., 2001; Yoo et al., 2001) mostly observed
bilateral (or sometimes left-dominant) top-down activation patterns within the auditory
cortex.
These studies employed tasks with overt meta-cognitive judgments (explicit
imagery with or without a visual cue, shadowing, instructed rehearsing, etc.) which might
recruit the left hemisphere more than would be the case in a less explicit experimental
setting. The present experiment, by contrast, indicated a clearly right-lateralised locus of
effect for the direct versus indirect speech comparison (Figure 2B) using a task that did not
involve any meta-cognitive judgments.
To account for this kind of spontaneous auditory imagery during silent reading, as well as
its modulation through linguistic reporting style (direct versus indirect speech), one might
adopt the notion of mental simulation in language comprehension, as proposed by
embodied cognition theories (e.g., Barsalou, 1999, 2008). Such theories argue that mental
representations of language are grounded in perceptual experiences and actions, and that
mental simulation (i.e., the mental re-enactment of perceptual, motor and introspective
states acquired during experience with the world, body and mind; Barsalou, 2009) is an
automatic and integral part of language comprehension.
Page | 43
This raises the question of (1) the nature of the perceptual experiences that underlie voice
simulation in response to reading direct versus indirect speech statements in text, and (2)
the nature of the representations that are activated during voice simulation.
Regarding the first question, we assume that accumulated experience with direct versus
indirect speech usage form the basis for voice simulation during silent reading.
As
discussed in the introduction, when speakers employ direct speech, they often mimic or
dramatise aspects of the reported speaker‟s voice in order to demonstrate or depict the
reported speech act; indirect speech, by contrast, is typically not used in such a vivid,
demonstrative fashion because its pragmatic function is to provide descriptions of what
was said (Clark & Gerrig, 1990). Comprehension of direct speech is therefore more likely
to be grounded in the perceptual experience of vivid vocal depictions of the reported
speech acts. This would explain why silent reading of direct speech is more likely to
engender mental simulations of voice than silent reading of indirect speech.
The second question concerning the exact nature of the simulated voice representations is
more difficult to answer at present. We conjecture that the simulated voice representations,
as well as their neural correlates, overlap to a large degree with those activated during
„encouraged‟ auditory imagery (see earlier discussion).
However, compared to the
speaker-specific voices that are likely to be activated during auditory imagery tasks
(particularly following pre-exposure to concrete speech samples), the simulated voice
representations reflected in the present study may be less specific, i.e., they may only
involve speaker-unspecific aspects of voice. The reason for this assumption is that our
experimental setup did not encourage the imagination of speaker-specific voices. One of
those speaker-unspecific aspects of voice that underlie the present voice simulation
findings could be emotional prosody (supra-segmental acoustic information characterising
emotional information), which has not only been found to be associated with a rightlateralised activation pattern (cf. Mitchell, Elliott, Barry, Cruttenden, & Woodruff, 2003;
Wiethoff et al., 2008; Wildgruber, Ackermann, Kreifelts, & Ethofer, 2006), but would also
fit well with the notion of direct speech as vivid demonstration (the reported speaker‟s
emotional state is often demonstrated via voice mimicking in direct speech reporting).
That said, the exact nature of the voice representations that are spontaneously activated
during silent reading of direct versus indirect speech still remains an important question for
future research. Part of this has been explored in Chapter 3.
Page | 44
From a broader perspective, it is worth noting that the right STS area (not restricted to the
clusters that we observed in the current experiment) is also partially involved in other
cognitive processes such as biological motion processing (Sokolov et al., 2011), dynamic
face processing (Pitcher, Dilks, Saxe, Triantafyllou, & Kanwisher, 2011), judging the
trustworthiness of faces (Dzhelyova, Ellison, & Atkinson, 2011), categorising spectrally
complex sounds (Klein & Zatorre, 2011), intention processing in communication (Enrici,
Adenzato, Cappa, Bara, & Tettamanti, 2011), perspective taking in Theory of Mind (ToM)
tasks (Dodell-Feder, Koster-Hale, Bedny, & Saxe, 2011) and eye gaze processing in social
interactions (Hooker et al., 2003). These findings indicate that the right STS area plays an
important role in human communication across different sensory and motor modalities.
In conclusion, the present experiment showed that even without being explicitly
encouraged to imagine voices, readers are more likely to mentally simulate, or
spontaneously imagine, aspects of the reported speaker‟s voice during silent reading of
direct speech as opposed to meaning-equivalent indirect speech.
The results can be
interpreted in line with embodied cognition (e.g., Barsalou, 1999) and the notion that direct
speech is represented in a more vivid and perceptually engaging fashion than indirect
speech (Clark & Gerrig, 1990). The present study is novel and important in several
respects. Indeed, it is the first demonstration that top-down activation of voice-sensitive
areas in the auditory cortex (Belin et al., 2000) can be modulated by pragmatically
different reporting styles. Second, it further examines on voice simulations during silent
reading. Our results are consistent with the embodied cognition hypothesis, extending it to
the auditory perceptual modality which so far has received little attention in the relevant
literature (except for studies that focused on sound-related words, see Kellenbach et al.,
2001; Kiefer et al., 2008). Third, this study combined event-related fMRI with eyetracking to investigate neural correlates during on-line reading of text under relatively
natural presentation conditions (contrasting with word-by-word reading, for example); this
could inspire new applications in psycholinguistic research on language comprehension,
helping us to understand the kinds of mental representations activated during reading.
Finally, it sheds new light on the distinction between direct versus indirect speech from a
cognitive neuroscience perspective, suggesting that perceptual vividness is one of the key
aspects differentiating the two.
Page | 45
Chapter 3 Brain ‘talks over’ boring quotes: Top-down activations
of voice-selective areas when listening to monotonous direct
speech utterances (Experiment 2)
3.1
Introduction
Combining event-related fMRI and eye-tracking, we found that silent reading of direct
speech elicited higher brain activity in voice-selective areas (Belin et al., 2000) of the right
auditory cortex than silent reading of indirect speech (Chapter 2). The findings are in line
with the demonstration theory of direct speech (Clark & Gerrig, 1990) and the embodied
theories of language processing (e.g., Barsalou, 1999; Zwaan, 2004), suggesting that
comprehension of direct speech is grounded in the experiences of more vivid vocal
depictions and that individuals are more likely to mentally simulate reported speakers‟
voices during silent reading of direct speech as opposed to indirect speech.
However, with no acoustic stimulation as a reference, it is still unclear what constitutes
such „inner voice‟ experiences during silent reading of direct as opposed to indirect speech.
In Chapter 2, we speculated that the mentally simulated voice representations entail suprasegmental acoustic information of the quoted speaker‟s voice (e.g., speech melodies,
intonation and emotional prosody), given that a right-lateralised activation pattern was
observed.
Indeed, „bottom-up‟ auditory stimulation studies have shown the same
lateralisation in comparisons of (1) speech or music with acoustically matched noise bursts
(Zatorre, Evans, & Meyer, 1994; Zatorre, Evans, Meyer, & Gjedde, 1992), (2) speech
signals (irrespective of intelligibility) with noise-vocoded signals (Scott, Blank, Rosen, &
Wise, 2000), and (3) nonverbal sounds comprising extended frequency transitions (suprasegmental) with those comprising rapid frequency transitions (sub-segmental) (Johnsrude,
Penhune, & Zatorre, 2000).
Hence, it seems likely that the right superior temporal
gyrus/sulcus (STG/STS) areas are involved in processing dynamic pitch variations which
are an important property of supra-segmental vocal information.
One type of such
information, namely emotional prosody and intonation, is also found to activate similar,
right-lateralised activation patterns in various forms including sentences (Mitchell, Elliott,
Barry, Cruttenden, & Woodruff, 2003; Wildgruber et al., 2005), words (Wiethoff et al.,
2008) and word-like vocalisations (Grandjean et al., 2005). Most importantly, mental
simulations of supra-segmental acoustic information in language comprehension would fit
Page | 46
well with the notion of direct speech as vivid demonstration - in which vivid depictions of
the quoted speaker‟s voice are characterised in terms of enriched supra-segmental acoustic
information.
In this chapter, we attempted to address whether the mentally simulated voices
predominantly consist of supra-segmental acoustic information during comprehension of
direct as opposed to indirect speech. In order to create acoustic references to verify such
supra-segmental information, we prepared audio recordings of short stories in which direct
and indirect speech utterances were spoken monotonously. This manipulation preserved
(sub)-segmental acoustic information such as the phonological information of the uttered
words in the recordings, but minimised supra-segmental acoustic information such as the
global intonation patterns over the utterances. Thus, if direct speech is represented in
enriched
supra-segmental
acoustic
representations
of
voices
during
language
comprehension, individuals would have to mentally simulate such representations, since
they are minimised in the stimuli, to supplement what they hear while listening to
monotonously spoken direct speech utterances.
By contrast, indirect speech is not
represented in vivid voice representations and hence individuals need not simulate suprasegmental acoustic information when listening to monotonous indirect speech utterances.
Thus, we predict that listening to monotonous direct speech quotations would elicit higher
„top-down‟ brain activity in voice-selective areas of the right auditory cortex (i.e., similar
to the brain areas identified in Chapter 2) than listening to equally monotonous, meaningequivalent indirect speech quotations.
To test this hypothesis, we employed functional magnetic resonance imaging (fMRI) to
measure participants‟ brain activity while they were listening to short stories which
contained monotonously spoken direct or indirect speech utterances. Between the two
conditions, we compared the evoked BOLD signal changes within participants‟ voiceselective brain areas. Moreover, we performed multiple parametric modulation analyses to
verify the underlying source of any differential brain activity we observed. To assess the
between-study consistency, we also compared the observed activation patterns with our
previous silent reading data (Chapter 2).
Page | 47
3.2
Method
3.2.1 Participants
Twenty-one adult participants were recruited and scanned. They were native English
speakers with normal hearing and language abilities, and with no history of neurological or
psychiatric disorders. Three participants had to be excluded from analysis due to either (a)
no clear response in the voice localiser task (1 subject) and/or excessive head-movements
during scanning (2 subjects), or (b) scanning abortion following claustrophobic symptoms
(1 subject). Data from the remaining 18 participants (age 18 – 32 years, 9 males and 9
females) were valid for the final analyses. All of them were right-handed except for one
female subject.
They signed the consent form and were paid at £6/hour for their
participation.
3.2.2 Stimuli
Listening stimuli
Ninety short stories with different protagonists (indicated by different names) were
recorded as stimuli.
The stories were exactly the same as in Chapter 2.
The full
transcriptions are listed in Appendix 1. Each story started with two declarative sentences
to set up a scenario (e.g., Luke and his friends were watching a movie at the cinema. Luke
wasn‟t particularly keen on romantic comedies, and he was complaining a lot after the
film.), followed by either a direct speech or an indirect speech quotation sentence (e.g., He
said: „God, that movie was terrible! I‟ve never been so bored in my life.‟ or He said that
the movie was terrible and that he had never been so bored in his life.). The reported
clauses in both conditions (underscored in the above examples) were equivalent in terms of
linguistic content. Additional comprehension questions were also recorded for 23 stories
(ca. 25%) to assess participants‟ overall comprehension accuracy and to ensure that they
listen to the stories for comprehension.
The stories and questions were spoken by a professional actress.
Critically, in one
condition, the direct speech utterances were deliberately spoken as monotonously (Directmonotonous condition) as the indirect speech utterances (Indirect condition), that is,
without providing vivid depictions of the reported speaker‟s voice. We also recorded
Page | 48
„normal‟ (i.e., vivid) versions of the direct speech utterances which were used as a control
condition
(Direct-vivid
condition).
Example
recordings
are
available
at:
[www.psy.gla.ac.uk/~boy/fMRI/samplerecordings/].
Three lists of stimuli with counterbalanced item-condition combinations (i.e., 30 Directmonotonous trials, 30 Indirect trials, and 30 Direct-vivid trials per list) were constructed
using a Latin square. Each item appeared once per list, but in a different condition across
lists. Each list was assigned to one third of our participants. The presentation order of the
items per list was randomised for each participant.
Voice localiser stimuli
For the voice localiser session (see 3.2.3 Procedure), we presented blocks of vocal sounds
and
non-vocal
sounds
provided
by
(vnl.psy.gla.ac.uk), University of Glasgow.
the
Voice
Neurocognition
Laboratory
These stimuli were the same as those
employed in Belin et al. (2000), and comprised both speech (e.g., spoken vowels) and nonspeech (e.g., laughing and coughing) vocal sound clips, as well as non-vocal sound clips
(e.g., telephone ringing and dog barking). The contrast in brain activity elicited by vocal
versus non-vocal sounds reliably localises voice-selective areas of the auditory cortex.
3.2.3 Procedure
Participants were positioned in the scanner, wearing MRI-compatible, electrostatic
headphones (NordicNeuroLab, Norway) for (1) auditory presentation during both the story
listening session and the voice localiser session and (2) noise attenuation during fMRI
scanning. For the story listening session, participants were instructed to keep their eyes
closed, to listen to the stories carefully and to answer comprehension questions which
would follow 25% of the short stories they had heard. The stimuli were presented using Eprime 2.0 (Psychology Software Tools, Inc., USA); each trial started with a 4-second
silence period, followed by the presentation of the story and then (on 25% of the trials) a
comprehension question regarding the content of the preceding story. Each such question
appeared one second after termination of the preceding story presentation and prompted a
„yes‟ or „no‟ response which participants could provide by pressing buttons on a response
box with their index or middle fingers, respectively. The 90 listening trials were evenly
Page | 49
interspersed with five 30-second „baseline‟ trials during which no experimental stimulation
was present.
After the story listening session, an anatomical scan of the participant‟s brain was
performed, followed by a brief (ca. 10-min) voice localiser scanning session. During the
latter, participants were instructed to close their eyes while listening to 20 8-sec blocks of
vocal and 20 8-sec blocks of non-vocal auditory stimuli presented in an efficiency
optimised, pseudo random order along with 20 8-sec blocks without stimulation, acting as
a baseline (cf. Belin et al., 2000).
3.2.4 MRI acquisition
Scanning was performed on a 3-T Siemens Tim Trio MRI scanner using a 12-channel head
coil (Erlangen, Germany). Functional scans (for both the story listening session and voice
localiser session) were acquired using a T2*-weighted echoplanar imaging (EPI) sequence
(32 slices acquired in orientation of the Sylvian fissure; TR = 2s; TE = 30 ms; matrix size:
70 × 70; voxel size: 3 × 3 × 3 mm; FOV = 210). T1 whole-brain anatomical scans were
obtained using 3D T1-weighted magnetization prepared rapid acquisition gradient echo
(MP-RAGE) sequence (192 axial slices; matrix size: 256 × 256; voxel size: 1 × 1 × 1 mm;
FOV = 256). The average scanning time for the whole experiment was around 55 min per
participant.
3.2.5 Data analysis
3.2.5.1
Whole brain and ROI analyses
All MRI data were analysed using SPM8 (www.fil.ion.ucl.ac.uk/spm/, University College
London). Pre-processing of functional scans included (a) head motion corrections (trilinear interpolation) whereby scans were realigned to the first volume; (b) co-registration
of functional scans to their corresponding individual anatomical scans; (c) segmentation of
the co-registered scans; (d) normalisation of functional (3 mm isotropic voxels) and
anatomical (1 mm isotropic voxels) data to Montreal Neurological Institute (MNI) space;
(e) smoothing of normalised data (10-mm Gaussian kernel).
Page | 50
fMRI data from the anatomical and voice localiser scanning sessions were used to
determine the voice-selective areas in the auditory cortex (ROI). The individual voice
localisers for most participants (12 subjects) were obtained at p < 0.05 (FWE-corrected).
The voice localisers for the other 6 subjects were obtained at p < 0.001 (uncorrelated, to
increase sensitivity). The group voice localiser was obtained at p < 0.05 (FWE-corrected).
For the story listening session, the temporal onset of a critical fMRI event was defined as
the temporal onset of the first word within the quotation marks (direct speech) or the
complementiser that (indirect speech); its offset was defined as the temporal offset of the
last word in the direct or indirect speech quotations. Non-critical events (e.g., listening to
background sentences, comprehension questions and instructions, as well as button
pressing) were specified as events of no interest in the design matrix. The rest consisted of
all „silence‟ events (including five 30-second baseline trials and all 4-second pre-trial
silence periods) and were regarded as baseline. The fMRI data were mapped to the human
Colin atlas surface (D. C. Van Essen, 2002) in CARET (D. C. Van Essen et al., 2001). The
mean beta estimates within ROIs were calculated by SPM toolbox easyROI
(www.sbirc.ed.ac.uk/cyril/cp_download.html), and submitted to 2-tailed paired-samples ttests.
3.2.5.2
Parametric modulation analyses
To verify the underlying source of the observed brain activations, we performed parametric
modulation analyses with (1) the acoustic parameters, (2) the perceived vividness and (3)
the perceived contextual congruency of the critical direct and indirect speech utterances.
The acoustic parameters were intended to objectively capture the acoustic characteristics of
the critical audio recordings independent of linguistic content. The parametric modulations
with these measures would unveil whether the observed brain activations were simply
engendered by the acoustic differences between conditions. Comparably, the vividness
ratings (see below) were intended to provide a more subjective measurement of the
acoustic characteristics of the critical direct and indirect speech utterances independent of
linguistic context. In a way, this measure summarises the joint effect of the acoustic
characteristics (i.e., vocal features) and linguistic content (wording etc.) on how vivid the
critical utterances would sound to perceivers irrespective of context. Thus, the parametric
modulations with the vividness measure would reveal whether between-condition
Page | 51
differences reflect evoked brain responses to the differential vocal features that are
subjectively perceived „bottom-up‟ from the stimuli. Finally, the contextual congruency
ratings were intended to measure the discrepancy between the actual vocal features of the
critical stimuli and the way these stimuli „should have sounded like‟ in the given contexts.
In other words, instead of quantifying the „bottom-up‟ perceived vocal vividness of the
stimuli, the contextual congruency metric was intended to capture the contextually
expected vividness (or its discrepancy with the actually perceived vividness) of the critical
speech stimuli.
In this sense, parametric modulation analyses with the contextual
congruency metric would indicate whether observed brain activation patterns reflect „topdown‟ mental simulations of enriched vocal depictions while listening to the critical
(monotonous) direct and indirect speech utterances.
Acoustic parameters
Using Praat software (Boersma & Weenink, 2010), we characterised the critical audio
samples in terms of eight acoustic parameters which are known to be related to speech
prosody: (1) the mean pitch (fundamental frequency F0 averaged over the duration of each
sample; (2) the pitch variation, measured as the standard deviation in pitch over the
duration of each sample (pitch SD); (3) the pitch range (difference between maximum and
minimum pitch in Hz over the duration of each sample); (4) the mean intensity (in dB) over
the duration of each sample; (5) the intensity variation, measured as the standard deviation
in intensity over the duration of each sample (intensity SD); (6) the intensity range; (7) the
duration of the voiced sample (the recording periods in which the pitch value passed the
voicing threshold); (8) the duration of the entire audio sample. These eight parameters
were then included simultaneously as modulators in the parametric modulation analyses to
partial out their joint contribution to the between-condition differences in the evoked
BOLD signals6.
6
Since we are solely concerned with the acoustic parameters‟ joint contribution to the brain activation
patterns of interest (and not with each parameter‟s individual importance), multicollinearity is not an issue
here.
Page | 52
Vividness ratings
We recruited 12 native speakers of English with normal hearing and language abilities for
this rating study. They were paid at £2 for their participation. A typical session took 10 –
20 min.
Participants were seated in front of a Dell Duo Core PC, wearing headphones. They were
presented with the same auditory stimuli that were used in fMRI scanning; only the critical
direct and indirect speech utterances (i.e., without context) were presented. After hearing
each utterance, participants had to give a rating by pressing number keys on a keyboard, to
indicate how vivid and engaging the utterance they had just heard was. The ratings were
given in a 7-point scale in which 7 meant „very vivid and engaging‟ while 1 meant
„extremely monotonous‟.
The collected ratings were summarised by condition for each subject and then submitted to
paired-sample t-tests to assess the between-condition differences in vividness. They were
also summarised by trial and were then included as a parametric modulator to partial out
the contribution of perceived vividness to the between-condition differences in the evoked
BOLD signals.
Contextual congruency ratings
We recruited a different set of 12 native speakers of English with normal hearing and
language abilities for this rating study. They were paid at £4 for their participation. A
typical session took 30 – 40 min.
The procedure of this rating study was the same as in the above vividness rating study
except: (1) participants were presented with the whole stories (i.e., with context), and (2)
they had to give 7 point-scale ratings on the „contextual congruency‟ of the critical direct
and indirect speech utterances, i.e., on whether these utterances matched the context in
terms of how vivid and engaging they were - 7 meant „fits in the context extremely well‟
while 1 meant „does not fit in the context at all‟.
The collected ratings were first summarised by condition for each subject and were then
submitted to paired-sample t-tests to assess the between-condition differences in contextual
Page | 53
congruency. They were also summarised by trial and were then included as a parametric
modulator to partial out the contribution of contextual congruency to the betweencondition differences in the evoked BOLD signals.
Parametric modulations
We performed three parametric modulation analyses with the speech utterances‟ acoustics,
vividness and contextual congruency as the modulators, respectively. This was to assess
each set of modulators‟ individual contributions to the observed brain-activation
differences between the Direct-monotonous condition and the Indirect condition. Our
strategy was to examine whether the differential brain activations would be reduced (i.e.,
accounted for) as a consequence of partialling out the effects of the investigated
modulators. First, we performed parametric modulations at the individual level. For each
participant, we specified in the design matrix a single trial-type for all three conditions
(including the control condition Direct-vivid); it was first followed by one of the three sets
of modulator(s) – this would ensure that the effects of the investigated modulator(s) are
partialled out across all trials – which were then followed by three experimental conditions
coded with binary values. After the experimental trials, other event types and participants‟
head motion parameters were also included in the design matrix for modelling the
collected BOLD signals. We then conducted the contrast analyses in the same way as
before. Using the same threshold (i.e., p < 0.05, FWE-corrected), we examined how the
observed brain-activation differences between the Direct-monotonous condition and the
Indirect condition were affected when the effects of each set of modulators were partialled
out, respectively.
3.2.5.3
Comparing brain activation patterns with Chapter 2
Moreover, we also compared the brain activation patterns observed in the current study (in
the contrast of Direct-monotonous against Indirect conditions) with the previous silent
reading study (Chapter 2). The comparison was descriptive: The activation patterns were
described using the 3D coordinates (in relation to the standard brain space from the
Montreal Neurological Institute) of the peak voxel within each activation „blob‟. We
paired the activation „blobs‟ with their counterparts between the two studies and compared
the peak voxels‟ coordinates within each pair.
Page | 54
3.3
Results and discussion
3.3.1 Whole brain and ROI analyses
The voice localiser
Consistent with the findings of Belin et al. (2000), we found that the vocal sounds elicited
significantly (ts > 7.6, ps < .02, FWE-corrected) greater activity than non-vocal sounds
bilaterally in the STG/STS areas (Figure 3A). The maximum of voice-sensitive activation
was located along the upper bank of the central part of the right STS.
The main contrast
Within the right voice-selective area, we found that listening to monotonously spoken
direct speech utterances elicited greater BOLD signals than listening to monotonous
indirect speech utterances (see Figure 3B). The differential brain activity between the
Direct-monotonous and the Indirect conditions were located in voice-selective areas along
the posterior, middle and anterior parts of the right STS brain areas. As a whole, no region
showed an opposite pattern of activity, and the Direct-monotonous condition was always
associated with a greater BOLD signal than the Indirect condition. Both conditions were
active against baseline. The between-condition difference was significant: For individual
Regions of Interest (ROIs), 2-tailed paired-sample t(17) = 5.650, p < 0.001; for the group
ROI, t(17) = 4.979, p < 0.001. The two monotonous conditions elicited significantly less
BOLD signal in the right voice-selective area than the Direct-vivid condition (Figure 3C).
Page | 55
Figure 3 Illustrations of the localizer and the between-condition differences (Direct-monotonous
against Indirect), A, Brain regions that selectively responded to vocal sounds as opposed to non-vocal
sounds (i.e., the voice-selective areas), B, Within the voice-selective areas (indicated by black lines), brain
regions that selectively responded to auditory comprehension of monotonous direct speech utterances as
opposed to monotonous indirect speech utterances in the whole brain analysis (n = 18), C, Mean signal
change (against baseline) between conditions in the right voice area Region of Interest (ROIs), determined
individually (left) or using the sample average (right). The single error bar in each panel refers to the 95% CI
for the between-condition difference (Direct-monotonous against Indirect only).
Page | 56
The higher brain activity in the Direct-vivid condition than in the two monotonous
conditions can be explained by the more intense and varied acoustic information that is
carried in the more vivid speech utterances (also suggested in Table 2 below). The higher
brain activity in the Direct-monotonous condition than in the Indirect condition could be
explained in two very different ways. First, it is possible that the Direct-monotonous
condition was somehow acoustically more varied than the Indirect condition. In this case,
the higher BOLD signals for the Direct-monotonous condition would simply be a
reflection of the more varied acoustic information that is carried in those stimuli. However,
it is also possible that the Direct-monotonous condition was acoustically equivalent to (i.e.,
not more varied than) the Indirect condition. In that case, the higher BOLD signals for the
Direct-monotonous condition must have been due to something other than „bottom-up‟
acoustic characteristics of the stimuli, making an explanation more likely whereby
monotonous direct speech utterances were supplemented „top-down‟ (via mental
simulations) with enriched supra-segmental acoustic information.
To distinguish between these contrasting interpretations, we conducted parametric
modulation analyses to verify the underlying source of the observed brain activations
between Direct-monotonous and Indirect conditions.
We investigated the signal
contribution of the speech utterances‟ acoustic parameters, their vividness and their
contextual congruency, respectively.
3.3.2 Parametric modulation analyses
It was found that the examined parametric modulators, i.e., the acoustic parameters, the
perceived vividness and the perceived contextual congruency, were highly correlated with
one another (Table 1). This is not surprising since they all measured variables associated
with the vocal vividness of the direct and indirect speech utterances, either objectively,
subjectively, or subjectively under consideration of linguistic context (see 3.2.5.2).
However, the correlations were certainly not perfect, and the parametric modulation results
below revealed rather distinct parametric contributions to the brain activation patterns of
interest.
Page | 57
Pearson
Correlation
(N = 270)
Pitch Mean
Pitch SD
Pitch Range
Loudness
Mean
Loudness SD
Loudness
Range
Voiced
Sample
Duration
Recording
Duration
Vividness
Rating
Contextual
Congruency
Rating
Pitch
Mean
Pitch
SD
Pitch
Range
Loudness
Mean
Loudness
SD
Loudness
Range
1
.748
.745
.748
1
.912
.745
.912
1
.715
.404
.402
.479
.398
.359
.592
.431
.385
Voiced
Sample
Duration
.156
.096
.201
.715
.404
.402
1
.528
.782
.479
.398
.359
.528
1
.699
.592
.431
.385
.782
.699
.156
.096
.201
.267
.162
.172
.270
.749
.713
.495
.521
Recording
Duration
Vividness
Rating
Contextual
Congruency
.162
.172
.270
.749
.713
.693
.495
.521
.509
.267
.161
.600
.297
.031
.138
.432
.261
1
.270
.299
.579
.254
.031
.270
1
.856
.232
.135
.161
.138
.299
.856
1
.288
.224
.693
.600
.432
.579
.232
.288
1
.632
.509
.297
.261
.254
.135
.224
.632
1
Table 1 The cross-correlations between the examined parametric modulators.
3.3.2.1
The effects of acoustics
The descriptives of the eight acoustic parameters are summarised by condition in Table 2.
It was found that some of the acoustic measures predict decreased STS activations in the
contrast of the Direct-monotonous condition to the Indirect condition. For instance, the
Direct-monotonous condition was associated with lower Pitch SD (therefore less varied
acoustic information) as compared with the Indirect condition, and should have elicited
less brain activity in the voice-selective areas. Therefore, these acoustic measures could
not explain the enhanced STS activations that we observed in the Direct-monotonous
condition as opposed to the Indirect condition. However, acoustic measures related to
intensity (i.e., Intensity Mean, Intensity SD and Intensity Range) were found to display a
congruent pattern with the observed brain activation difference between the two conditions:
The Direct-monotonous condition was associated with significant higher Intensity Mean,
SD and Range, paired-sample ts(89) > 2.5, ps < .013. Although the between-condition
differences in these acoustic measures were statistically significant, the sizes of the
differences do not seem to be evidently meaningful in practice. For instance, the Intensity
Mean difference between the Direct-monotonous and the Indirect conditions amounted to a
mere 1.12 dB, which is hardly noticeable in loudness perception, given that a 5 dB
increment step size is usually recommended for subjects with normal-hearing sensitivity
(Cox, Alexander, Taylor, & Gray, 1997). Nonetheless, if these intensity differences did
contribute to the brain-activation differences between the two critical conditions, we expect
that by partialling out the effects of these acoustic measures in the following parametric
Page | 58
modulation analysis, the between-condition differences in brain activations should be
subsequently accounted for (i.e., reduced).
Condition
Direct-monotonous
Indirect
Direct-vivid
200.43 (12.42)
208.50 (13.39)
265.81 (43.22)
Pitch Mean (Hz)
32.62 (6.97)
45.24 (10.51)
66.36 (20.11)
Pitch SD (Hz)
143.87 (39.27)
199.97 (52.80)
281.77 (82.48)
Pitch Range (Hz)
68.01 (1.17)
66.89 (.93)
70.71 (2.48)
Intensity Mean (dB)
7 (.51)
6.87 (.44)
7.51 (.68)
Intensity SD (dB)
31.23 (1.62)
29.66 (1.59)
34.32 (3)
Intensity Range (dB)
2044
(534)
2073
(559)
2318
(662)
Voiced Duration (ms)
4028
(984)
4332
(1056)
4702
(1119)
Audio Sample Duration (ms)
Table 2 A summary of the eight acoustic parameters (means, with standard deviations in parentheses)
for the spoken stimuli in each experimental condition.
Acoustic parameters
The parametric modulation analyses showed that, using the same threshold (p < 0.05,
FWE-corrected), the Direct-monotonous condition still elicited higher brain activity within
the right voice-selective area than the Indirect condition after the effects of the acoustic
parameters were partialled out (Figure 6B). Since the activation patterns were highly
similar to the original contrast (Figure 6A), it suggests that the observed brain-activation
differences between the Direct-monotonous and the Indirect conditions were unlikely to be
engendered by the between-condition differences in acoustic characteristics.
3.3.2.2
Partialling out the effect of vividness
The averaged ratings and the paired-sample t-test results are illustrated in Figure 4. It was
found that the control condition Direct-vivid was perceived as significantly more vivid
than the two monotonous conditions, ts(11) > 19, ps < .001. However, instead of being
„equally monotonous‟, the Direct-monotonous condition was perceived as significantly less
vivid than the Indirect condition, paired-sample t(11) = -7.872, p < .001. The latter
suggests that the monotonous direct speech utterances (Direct-monotonous) contained less
vivid vocal modulations than the monotonous indirect speech utterances (Indirect). This is
also inconsistent with a more „bottom-up‟ explanation of our findings (Figure 3): Given
lower vividness ratings for the Direct-monotonous condition, a „bottom-up‟ account would
predict that this condition would consume less energy (therefore less blood oxygen) to
process within the voice-selective areas than the more vivid Indirect condition; in other
words, the Direct-monotonous condition should have elicited significantly decreased
Page | 59
BOLD signals within these brain areas compared to the Indirect condition. However,
exactly the opposite was found (see 3.3.1).
Figure 4 Illustrations of the between-conditions differences in vividness. The significance of the
pairwise contrasts (referred to with square brackets) were indicated with asterisks (*** indicates p < .001).
The single error bar represents the 95% CI of the between-condition difference between Direct-monotonous
and Indirect conditions.
The parametric modulation analysis on vividness revealed that using the same threshold (p
< 0.05, FWE-corrected), the Direct-monotonous condition still elicited higher brain
activity within the right voice-selective area than the Indirect condition after the effects of
vividness were partialled out (Figure 6C); the between-condition difference appeared to be
larger and more widely spread than the original contrast (Figure 6A). This suggests that
the originally observed brain-activation difference between the Direct-monotonous and the
Indirect conditions is not „explained‟ by the vividness of the speech utterances (for such a
conclusion to be justified, there should have been a reduction in the original difference
after partialling out the effect of vividness). Instead, it indicates that the original betweencondition difference was partially „masked‟ by the vividness contrast between the two
conditions: The vividness ratings indicate that the Direct-monotonous condition was
perceived as less vivid than the Indirect condition (see Figure 3); when the two conditions
were brought to the same vividness level (by partialling out the effect of the vividness
modulator), the original brain-activation difference was enhanced. In other words, the
influence of factors other than vividness became more pronounced when the negative
contribution of vividness (Direct-monotonous < Indirect) was eliminated. We conclude
that while perceived vividness clearly played a role in the originally reported brain
Page | 60
activation patterns, its contributions actually went contrary to an actual explanation of
those brain activation patterns.
3.3.2.3
Partialling out the effects of contextual congruency
The averaged ratings and the paired-sample t-test results are illustrated in Figure 5. It was
found that while the Indirect and the Direct-vivid conditions were perceived as equally
congruent with the preceding linguistic contexts, paired-sample t(11) = 1.754, p > .1, the
Direct-monotonous condition was perceived as significantly less congruent with contexts
than both of them, paired-sample ts(11) > 4, ps < .003 (Figure 5). The results show that
listeners routinely expect vivid vocal depictions for direct speech but not for indirect
speech; they have to mentally simulate or imagine vivid depictions of the reported
speaker‟s voice to supplement the monotonously spoken direct speech utterances
(contextually incongruent) but not the monotonous indirect speech utterances or the vivid
direct speech utterances (contextually congruent). This voice simulation process in the
Direct-monotonous condition would have required additional energy consumption within
the voice-selective areas of the auditory cortex; this could explain why the less vivid
Direct-monotonous condition (see 3.3.2.2), which should have elicited decreased BOLD
signals in the voice-selective areas of the auditory cortex under a more „bottom-up‟
interpretation, actually elicited significantly higher BOLD signals within these brain areas
as compared to the Indirect condition.
Figure 5 Illustrations of the between-conditions differences in contextual congruency. The significance
of the pairwise contrasts (referred to with square brackets) were indicated with asterisks and abbreviations
(*** indicates p < .001, ** = p < .01, n.s. = not significant). The single error bar represents the 95% CI of
the between-condition difference between Direct-monotonous and Indirect conditions.
Page | 61
The parametric modulation analyses revealed that under the same threshold (p < 0.05,
FWE-corrected), only a few voxels at the right pSTS area survived when contrasting the
Direct-monotonous condition to the Indirect condition after the effects of contextual
congruency were partialled out (Figure 6D); compared to the original contrast (Figure 6A),
the between-condition difference had almost disappeared. The results confirm that the
observed between-condition difference reflects mental simulations of enriched suprasegmental vocal depictions specifically when listening to monotonously spoken direct
speech utterances rather than to monotonous indirect speech utterances.
3.3.2.4
Summary
The parametric modulation results showed that the increased brain activity during auditory
language comprehension of monotonous direct speech as opposed to monotonous indirect
speech can in part be explained by the contextual congruency of the direct or indirect
speech utterances, but not by their acoustic characteristics or their perceived vividness out
of context. It suggests that listeners routinely expect vivid depictions of the reported
speaker‟s voice for direct speech but not for indirect speech, and that they are more likely
to mentally simulate such enriched supra-segmental vocal representations while listening
to direct speech utterances which are spoken monotonously as opposed to monotonous,
meaning-equivalent indirect speech utterances.
Page | 62
Figure 6 Illustrations of between-condition differences (Direct-monotonous against Indirect) in
different parametric modulation models (the voice-selective areas are indicated by black lines), A, the
original contrast, B, the same contrast with effects of acoustics partialled out, C, the same contrast with
effects of vividness partialled out, D, the same contrast with effects of contextual congruency partialled out.
Page | 63
3.3.3 Comparing brain activation patterns between studies
It was found that the activation patterns observed when listening to monotonous direct
speech against monotonous indirect speech resembled those observed in silent reading of
direct speech against indirect speech (cf. Chapter 2). The brain activations patterns in both
studies were located at the posterior, the middle and the anterior parts of the right STS
areas. Within the MNI space, the peak voxels within each activation cluster were spatially
close to their counterparts across the two studies (Figure 7).
The between-study
consistency in the activation patterns suggests that the „inner voice‟ we observed in silent
reading of direct as well as indirect speech (Chapter 2) is similar in nature to the
enrichment of monotonous direct speech (as opposed to indirect speech) that we found in
the current study. Given that the „vocal enrichments‟ we observed in the present study
entailed supra-segmental acoustic information that was hardly available in the actual
stimuli, the „inner voice‟ we observed in silent reading may also have been suprasegmental in nature.
Figure 7 Illustrations of the critical contrasts between the two studies (masked by voice-selective
areas). The top panel shows the contrast between Direct-monotonous and Indirect conditions in the current
study (listening). The bottom panel shows the contrast between Direct speech and Indirect speech in
Chapter 2 (silent reading). The arrows point to the peak voxels‟ 3D coordinates (in MNI space) in the
activation clusters. The peak voxels were paired with their anatomical counterparts between the two studies.
The thresholds for the two contrasts were adjusted to better illustrate the activation blobs.
Page | 64
3.4
Summary and general discussion
The current experiment investigated mental simulations of supra-segmental acoustic
representations during auditory language comprehension of direct as opposed to indirect
speech. We employed audio recordings in which direct and indirect speech utterances
were spoken monotonously. This manipulation preserved the (sub)-segmental acoustic
information (e.g., phonological information associated with individual words) but
suppressed the supra-segmental acoustic information (e.g., the intonation patterns across
the speech utterances). Using event-related fMRI, we found that listening to monotonously
spoken direct speech utterances elicited significantly higher brain activity within voiceselective areas of the right auditory cortex as compared to listening to monotonous,
meaning-equivalent indirect speech utterances. Part of the between-condition difference
was significantly accounted for by the perceived contextual congruency of the direct and
indirect speech utterances but not by the utterances‟ acoustic characteristics or out-ofcontext vividness (see 3.2.5.2 for the rationale behind these three measures).
The results might be interpreted in terms of violations of listeners‟ expectation of vivid
vocal depictions for direct speech, which may not necessarily require mental simulations of
the expected supra-segmental acoustic information. However, fMRI studies shows that the
brain networks associated with violations of expectation typically consist of frontal cortex
such as dorsolateral prefrontal cortex (dlPFC) and ventrolateral PFC, lateral parietal cortex
and/or the basal ganglia but not the right STS area (Bubic, von Cramon, Jacobsen,
Schroger, & Schubotz, 2009; Huettel, Mack, & McCarthy, 2003). Moreover, the detection
of discrepancies/incongruencies between the semantic content (e.g., She won the lottery
jackpot) and the emotional prosody of utterances (e.g., a sad emotional prosody) has been
found to activate the left inferior frontal gyrus (Mitchell, 2006), as well as the right dorsal
anterior cingulate cortex (dACC) and the right mid STG/STS areas. These studies show
that violations of expectation are typically associated with more spreading brain activations
distributed among a network of brain areas. Importantly, the frontal brain areas seem to
play a critical role in processing conflicts between perception and expectation. Given that
in the current experiment, the brain-activation differences were only observed along the
right STS areas, the „expectation violation‟ account does not fit as well with the existing
evidence. One possibility may be that with no explicit incongruency detection task (the
current experiment employed a normal language comprehension task), the brain can
Page | 65
resolve the violations of expectation of vivid vocal depictions locally within the auditory
cortex without communicating with the frontal cortex. Further research is needed to verify
this account.
An alternative and perhaps better interpretation of the results adopts the notion of
perceptual simulations (e.g., Barsalou, 1999): Listeners expect vivid vocal depictions (i.e.,
enriched supra-segmental acoustic information) for direct speech but not for indirect
speech; they are more likely to mentally simulate such information (if it is not available in
the stimulus itself), thereby effectively supplementing what they hear when listening to
monotonously spoken direct speech utterances (contextually incongruent) but not
monotonously spoken indirect speech utterances (contextually congruent).
Unlike the
„expectation violation‟ account, this interpretation fits well, both theoretically and
empirically, with the previous findings on voice simulations in reading (Yao, Belin, &
Scheepers, 2011; Yao & Scheepers, 2011). For this reason, we are leaning towards a
„simulation‟ account in that the greater right STS activations when listening to monotonous
direct speech reflect enhanced mental simulations of vivid vocal depictions (i.e., enriched
supra-segmental acoustic information).
The current experiment sheds new light on the nature of the „inner voice‟ representations
that are mentally simulated in language comprehension of direct as opposed to indirect
speech. In the previous silent reading experiment (Chapter 2), there was no experimentally
manipulated auditory stimulation which could be used as a reference to the representational
nature of the observed „inner voice‟ activations. The current study resolved this issue by
using monotonously spoken direct versus indirect speech utterances in which the enriched
supra-segmental vocal information is scarcely available while the (sub)-segmental acoustic
information is intact. This manipulation suggested that the mentally simulated (as opposed
to acoustically perceived) voice representations must be supra-segmental rather than (sub)segmental in nature. Intriguingly, the „inner voice‟ activations observed in the current
investigation and in Chapter 2 were located in virtually the same brain areas. Reconciling
the findings of the two studies, we infer that the „inner voice‟ we observed in silent reading
of direct as opposed to indirect speech may also entail supra-segmental acoustic
representations.
Page | 66
At a broader level, the current findings also provide novel insights into speech perception.
Speech perception is not a passive information processing mechanism in that it involves
both bottom-up and top-down processes.
Various top-down influences in speech
perception have been documented: Listeners use prior lexical knowledge or perceptual
experience in perceptual grouping of speech, segmenting connected speech, perceptual
learning of distorted speech, and perceiving speech categorically (see Davis & Johnsrude,
2007 for a review). However, such top-down influences have mostly been documented in
terms of low-level „speech recognition‟ - researchers have mainly focused on how listeners
use prior lexical knowledge or perceptual experience to interpret distorted, unintelligible,
or ambiguous speech.
This neglects the fact that speech perception also involves
comprehension of the recognised linguistic information. It is currently less clear whether
and how on-line comprehension of auditory linguistic information also influences
upcoming intelligible and unambiguous speech. Moreover, many studies examine the topdown interactivity at the (sub)-segmental level (e.g., word interpretation), whereas the topdown influences at the supra-segmental level (e.g., emotional prosody) have received
limited attention.
The current study demonstrated top-down activations of supra-
segmental acoustic representations during intelligible and unambiguous speech perception
of direct versus indirect speech utterances. It provides evidence that during natural speech
perception, top-down interpretations of incoming acoustic signals routinely take place even
at the supra-segmental level.
More importantly, such top-down interpretations are
modulated as a function of linguistically/pragmatically different reporting styles (direct
versus indirect speech). Our findings emphasise that in addition to prior lexical knowledge
and perceptual experience, other linguistic factors such as reporting style should also be
considered in modelling the top-down interactivity of natural speech perception.
In conclusion, the current study shows that listeners routinely expect vivid depictions for
direct speech but not for indirect speech; they spontaneously engage in mental simulations
of vivid vocal depictions while listening to monotonously spoken direct speech as opposed
to monotonously spoken indirect speech. The findings replicate our previous findings of
an „inner voice‟ during silent reading of direct as opposed to indirect speech, but within a
different perceptual modality. This highlights the universality of such voice simulation
process in comprehension of direct speech. Furthermore, it provides evidence that the
nature of the mentally simulated „inner voice‟ entails supra-segmental acoustic
Page | 67
representations. It also verifies the neural correlates of such voice simulation process,
which include the anterior, the middle and the posterior parts of the right STS brain areas.
Future research should seek to specify the exact function of the brain areas involved in
such simulation processes. Finally, from a broader perspective, the current findings extend
the scope in modelling natural, intelligible speech perception, emphasising that
comprehension-driven, top-down influences at the supra-segmental level should also be
considered.
Page | 68
PART 2: The Acoustic Aspects of the ‘Inner Voice’ in
Comprehension of Direct versus Indirect Speech
Using event-related fMRI in both silent reading (Chapter 2) and listening (Chapter 3), we
have demonstrated that direct speech elicited higher brain activity in voice-selective areas
of the right auditory cortex as compared to indirect speech. The findings suggest that
individuals are more likely to mentally simulate enriched supra-segmental acoustic
representations of reported speakers‟ voices during language comprehension of direct
speech quotations.
However, it was still underspecified what acoustic information
constitutes such supra-segmental acoustic representations, and whether there would be
potential behavioural consequences of voice simulation during reading (e.g., modulation of
eye-movement patterns) and voice perception.
In the second stage of this doctoral research, we employed a range of behavioural
methodologies and attempted to address these questions by decomposing the purported
„inner voice‟ representations into acoustical parameters. Specifically, we focused on three
fundamental acoustic parameters of a speaker‟s voice, i.e., pitch (Chapter 4), loudness
(Chapter 5) and rate (Chapter 6), respectively. We examined how individuals‟ respond to
the direct and indirect speech reporting styles as well as the preceding linguistic contexts
during both oral and silent reading.
Page | 69
Chapter 4 Contextual modulation of pitch representations in oral
and silent reading of direct versus indirect speech
4.1
Introduction
The present chapter examined the pitch dimension of the observed „inner voice‟ experience,
namely the contextually implied pitch of the reported speaker, during language
comprehension of direct as opposed to indirect speech. Pitch (i.e., fundamental frequency,
F0) is an important acoustic parameter of a speaker‟s voice. It carries various types of
information about a speaker such as his/her gender (Kent & Read, 1992) and emotional
state (Banse & Scherer, 1996; Banziger & Scherer, 2005; Busso, Lee, & Narayanan, 2009).
It also signals the structures of speech utterances (e.g., their syntactic and prosodic
boundaries; see Beattie, Cutler, & Pearson, 1982; Danly, Cooper, & Shapiro, 1983; Hird &
Kirsner, 2002; Menn & Boyce, 1982; Umeda, 1981; Vaissiere, 1983) and what is
emphasised in a speech utterance (Pell, 2001). It is also a beneficial cue for decoding
speech in noisy environments (Brown & Bacon, 2010). Hence, given the importance of
pitch in verbal communication, we hypothesised that the observed „inner voice‟ experience
in comprehension of direct speech may contain representations of the reported speaker‟s
pitch.
One way of detecting such pitch representations is to examine whether individuals would
adjust their own pitch in accordance with the contextually (linguistically) implied pitch of
the reported speaker during oral reading of direct versus indirect speech. Oral reading
initially requires comprehension of the presented written text, which is followed by oral
interpretations of the formed linguistic representations into strings of vocal sounds (i.e.,
speech).
Since our brain-activation investigations (Chapter 2 & 3) have shown that
individuals are more likely to simulate reported speakers‟ voices during comprehension of
direct speech rather than indirect speech, we predict that individuals would be more likely
to engage in similar voice simulations during oral reading of direct speech as opposed to
indirect speech; critically, while the simulated voice representations are being orally
interpreted into speech, the acoustic characteristics of such representations may be
reflected in concordant modulations of vocal acoustics. In other words, if individuals
mentally simulate pitch information in accordance with the contextually implied pitch of
Page | 70
the reported speaker during oral reading, they may interpret such information by adjusting
their own pitch when reading direct speech, but less so when reading indirect speech.
Experiment 3 was aimed to test this hypothesis. We prepared short stories in which we
manipulated the reported speaker‟s pitch level by manipulating their emotional arousal and
gender. On one hand, it has been shown that a speaker‟s emotional arousal (Banse &
Scherer, 1996; Bulut & Narayanan, 2008) and gender (Kent & Read, 1992) are reliably
reflected in modulations of pitch (e.g., mean pitch,
0
0)
during articulation. For example,
is positively correlated with the speaker‟s emotional arousal: High-arousal emotion
such as Anger or Joy is generally associated with increased
emotion such as Sadness is associated with lowered
0
0
whereas a low-arousal
(Banse & Scherer, 1996). During
oral reading of direct speech rather than indirect speech, if individuals mentally simulate
reported speakers‟ pitch information, they may increase their own pitch when they read
about a reported speaker who is in a high-arousal emotion (e.g., Anger) as opposed to when
they read about a speaker in a low-arousal emotion (e.g., Sadness). On the other hand,
0
also indicates a speaker‟s gender: Women tend to speak in a higher pitch than men because
of different anatomical structures of their vocal tracts (Kent & Read, 1992). Hence, we
also predicted that individuals may increase their pitch during oral reading of direct rather
than indirect speech when the reported speaker is female (e.g., She said) as compared to
when the reported speaker is male (e.g., He said).
In addition to the above pitch manipulations, we also manipulated linguistic focus on
reported speakers‟ voices in order to verify the nature of the representations that underlie
any observed contextual modulation of
speech.
Linguistic
focus
0
modulates
during oral reading of direct versus indirect
mental
representations
during
language
comprehension (Sturt, Sanford, Stewart, & Dawydiak, 2004; Taylor & Zwaan, 2008). For
example, Taylor and Zwaan‟s previous findings (Zwaan & Taylor, 2006) demonstrated
that written descriptions of manual rotation (e.g., Eric turned down the volume) activate
compatible motor representations (i.e., an anticlockwise manual rotation) which interacted
with participants‟ concurrent manual rotation (for self-paced reading, using a knob). They
named such effects (i.e., the interactions between reading and manual rotation) motor
resonance. They found that motor resonance was localised to the verb that denoted the
action (i.e., turned down) and decayed quickly during reading of the post-verb information.
Page | 71
In their later study (Taylor & Zwaan, 2008), they explored whether mental simulations of
actions in reading were maintained while the actions were within linguistic focus and were
diminished as linguistic focus shifted away. They found that when a postverbal adverb
maintains linguistic focus on a matching action (e.g., „slowly‟ in relation to „turned down
the volume‟), motor resonance is accordingly maintained (i.e., mental representation of an
anticlockwise rotation is maintained); when a postverbal adverb shifts the linguistic focus
to the agent (e.g., „eagerly‟), motor resonance ceases. Their results suggest that motor
simulations are yield to linguistic focus. Thinking along the same lines, we postulated that
voice simulations may also be modulated by linguistic focus. The rationale here is that if
the purported voice simulations during oral reading of direct speech (should be reflected in
pitch modulations by reported speakers‟ emotional arousal and gender) can be selectively
modulated by exerting linguistic focus on voices (as opposed to non-vocal aspects such as
faces), we would have evidence that their underlying representations are likely to be voicerelated and to contain pitch information. Thus, in Experiment 3, we manipulated linguistic
focus on voices and investigated its effects on voice simulations during reading. Because
the critical direct speech and indirect speech quotations were always the final sentences in
our stimuli, we could not manipulate linguistic focus after quotations (i.e., the critical
regions for simulations) as Taylor and Zwaan had. As an alternative, we manipulated
linguistic focus before the critical quotations by using voice-focused (e.g., His voice was
furious and raspy: „…‟) versus voice-unfocused (e.g., He said, with his face turning red:
„…‟) linguistic descriptions. Assuming that exerting linguistic focus on reported speakers‟
voices has an effect on the subsequent oral reading of direct and indirect speech quotations.
In principle, it could have two contrasting effects. One possibility is that linguistic focus
on voices may enhance mental simulations of voices during oral reading of direct speech
but not indirect speech. In other words, the expected emotion effects (i.e., the emotional
arousal of the reported speakers) and gender effects (i.e., the gender of the reported
speakers) on participants‟ reading pitch may be more robust when a direct speech
quotation is preceded with a voice-focused description as opposed to when it is preceded
with a voice-unfocused description. Thus, we should observe that the pitch modulation
difference between oral reading of direct speech and indirect speech quotations would be
increased when the linguistic focus is exerted on reported speakers‟ voices (see Figure 8A).
The other possibility is that the linguistic focus on voices does not necessarily enhance
voice simulations in oral reading of direct speech because direct speech quotations alone
Page | 72
can trigger voice simulations sufficiently (a ceiling effect), but it enhances voice
simulations in oral reading of indirect speech (see Figure 8B). Thus, the difference in pitch
modulations between oral reading of direct speech and indirect speech quotations would
therefore be reduced or “masked” when linguistic focus is exerted on reported speakers‟
voices. While linguistic focus on voices is expected to modulate voice simulations in oral
reading of direct versus indirect speech, linguistic focus on non-vocal aspects (i.e., the
direct and indirect speech quotations are preceded with voice-unfocused descriptions)
should not affect the pitch modulations in oral reading of direct versus indirect speech. If
we observe such selective modulations (either enhancing or masking) of voice simulations
in oral reading by exerting the linguistic focus on voices, we could conclude that the
underlying representations of the observed pitch modulations during oral reading of direct
versus indirect speech are likely to be voice-related in nature and may contain pitch
information.
In sum, Experiment 3 investigated the pitch aspect in mental simulations of voices during
oral reading of direct versus indirect speech. We examined whether individuals adjust
their own pitch (
0)
in accordance with the reported speaker‟s emotional arousal and
gender that are contextually (linguistically) implied in the preceding context when reading
direct speech but not indirect speech. We also manipulated the linguistic focus on reported
speakers‟ voices to verify the voice-related nature of the underlying representations of any
observed
0
modulations during oral reading of direct versus indirect speech. Participants
orally read written stories which contained direct or indirect speech quotations while audio
recordings of their voices were taken for acoustic analyses.
Page | 73
Figure 8 Illustrations of the potential effects of linguistic focus on voices on voice simulations in oral
reading of direct versus indirect speech. A, the illustration of the „enhancing‟ effect, B, the illustration of
the „masking‟ effect.
The oral reading paradigm is useful in exploring acoustic parameters of the „inner voice‟
experience by examining individuals‟ articulation during oral reading of direct versus
indirect speech. However, it is also important to explore mental simulations of such
acoustic parameters during silent reading of direct versus indirect speech (i.e., without the
demand for articulation). One way of studying mental representations in silent reading is
to adopt the reaction time-based „reading & recognition‟ paradigms that have been
Page | 74
commonly used in the literature (e.g., Kurby et al., 2009; Stanfield & Zwaan, 2001; Zwaan
et al., 2002). The „reading & recognition‟ paradigms typically consist of a main reading
task and a secondary recognition task.
They examine the effects of language
comprehension (in the reading task) on subsequent recognition of the relevant perceptual
stimuli (in the recognition task): If the relevant perceptual representations (e.g., a
horizontal orientation of a nail) are mentally simulated during reading (e.g., of He
hammered the nail in the wall), they are expected to differentially influence individuals‟
recognition latencies (i.e., reaction times) to the perceptually-congruent (e.g., a picture
depicting a horizontally presented nail) versus -incongruent probe stimuli (e.g., a picture
depicting a vertically presented nail). Using similar paradigms in our investigations, we
predicted that if individuals mentally simulate reported speakers‟ pitch representations (e.g.,
a high-pitch voice of a woman) during silent reading of direct rather than indirect speech,
they have an advantage in recognising pitch-congruent stimuli (e.g., a high-pitch voice of a
woman) as opposed to pitch-incongruent stimuli (a low-pitch voice of a man).
We developed three versions (Experiments 4 - 6) of such „reading & recognition‟
paradigms to explore the hypothesised mental simulations of pitch during silent reading of
direct versus indirect speech. Because that we did observe noticeable interaction between
linguistic focus and the Reporting Style × Emotion interaction in Experiment 3 (i.e., the
underlying representations of the Emotion effects may not be voice-related), as well as for
the purpose of experimental design simplification, we exclusively focused on mental
simulations of pitch information in relation to reported speakers‟ genders in our silent
reading investigations. We examined the effects of genders on the recognition latencies to
a spoken probe word when its speaker‟s gender was made relevant by the previous text
passage. Participants silently read a series of short stories containing either a direct speech
or an indirect speech, each followed by a spoken probe word, and answered „yes‟ or „no‟ to
the spoken word on a secondary task. The probe words were spoken by either a male or a
female speaker such that it either matched or mismatched the reported speaker‟s gender
described in the preceding story.
Just as actually hearing a voice would facilitate
recognition of a congruent voice later, so too should simulating the voice facilitate later
recognition of a congruent voice. The greater the congruency between the simulated voice
and the heard voice (i.e., the more acoustic features they have in common), the greater the
degree of facilitation. If pitch representations are mentally simulated in accordance with
Page | 75
the reported speakers‟ genders during silent reading of direct as opposed to indirect speech,
recognition of probe words should be facilitated when the words are spoken by a „samegender‟ (i.e., same as the reported speaker‟s gender) speaker as compared to when they are
spoken by an „opposite-gender‟ speaker because the former are acoustically more
congruent with the simulated voices in pitch. Thus, we hypothesised that individuals
should respond faster to words that are spoken by a „same-gender‟ speaker than to words
spoken by an „opposite-gender‟ speaker immediately after or during silent reading of direct
speech but not indirect speech.
Experiment 4 employed a self-paced reading task,
followed by a spoken word recognition task in which participants had to indicate whether
the spoken word had been mentioned in the story they had just read. Experiment 5 utilised
the same self-paced reading task, which was, however, followed by a different lexical
decision task in which participants had to indicate whether the spoken word was a
meaningful English word or not.
The lexical decision task was believed to exert a
significantly lighter memory demand as compared to Experiment 4‟s recognition task7; it
was intended to retain the natural reading process (which does not require a heavy memory
load) as much as possible. While both Experiment 4 and 5 probed the simulated voice
representations off-line (i.e., the spoken words were presented 1000 ms after silent reading
of text), Experiment 6 employed an serial visual presentation (SVP) paradigm in which the
written critical words and the spoken words were presented simultaneously.
This
paradigm was designed to maximise the sensitivity in detecting voice representations. It
has been shown that mental simulations of motor representations during language
comprehension are dynamic and short-lived (Borreggine & Kaschak, 2006; Glenberg et al.,
2008; Zwaan & Taylor, 2006); hence it is possible that the simulated voice representations
are also short-lived and will decay quickly after reading. If this was the case, the off-line
paradigms used in Experiments 4 & 5 may not be sensitive enough to detect the simulated
voice representations – the latter may have already decayed before the presentation of
spoken probe words. By contrast, simultaneous presentation of the written critical words
(for reading) and the spoken probe words (for recognition) would be able to promptly
detect even the short-lived representations of voices.
7
The recognition task requires memorisation of every single word of a story. Such a heavy memory load
may disrupt normal reading process.
Page | 76
In sum, Experiments 4 to 6 investigated the pitch aspect in voice simulations during silent
reading of direct as opposed to indirect speech. We examined whether silent readers
would respond differentially to words that were spoken by a same-gender speaker (i.e., the
same as the reported speaker‟s gender) versus by an opposite-gender speaker. Participants
silently read short stories and responded to spoken words immediately after or during silent
reading of direct or indirect speech quotation in the story. Their response times and
accuracy were recorded for analysis.
4.2
Experiment 3
4.2.1 Participants
Forty-eight native speakers of English (sixteen male) with normal reading ability were
recruited for this experiment. They were paid £3 for their participation. A typical session
lasts about 25 min.
4.2.2 Stimuli and design
A 2 [Reporting Style: direct versus indirect speech] × 2 [Linguistic Focus: voice-focused
versus voice-unfocused pre-speech descriptions] × 2 [Gender: female versus male reported
speaker] × 3 [Emotion: high versus medium versus low emotional arousal of the reported
speaker] 8 design was employed (see examples below).
8
Emotion was a between-item manipulation that could not be counterbalanced within items. Hence the
examples for the high, medium and low arousal stories are presented separately from the examples in the
table below (in the following text).
Page | 77
Male reported speaker
Female reported speaker
Direct
speech
PhD student Ella was summoned to
her supervisor Jim’s office to give a
report on her current progress. Ella
asked for an extension but Jim looked
concerned. His voice turned serious:
‘Hmm…we really need those data in
by next month for that conference.’
PhD student Jim was summoned to his
supervisor Ella’s office to give a report
on his current progress. Jim asked for
an
extension
but
Ella
looked
concerned. Her voice turned serious:
‘Hmm…we really need those data in
by next month for that conference.’
Indirect
speech
PhD student Ella was summoned to
her supervisor Jim’s office to give a
report on her current progress. Ella
asked for an extension but Jim looked
concerned. His voice turned serious,
saying that they really needed those
data in by next month for that
conference.
PhD student Jim was summoned to his
supervisor Ella’s office to give a report
on his current progress. Jim asked for
an
extension
but
Ella
looked
concerned. Her voice turned serious,
saying that they really needed those
data in by next month for that
conference.
Direct
speech
PhD student Ella was summoned to
her supervisor Jim’s office to give a
report on her current progress. Ella
asked for an extension but Jim looked
concerned. His face turned serious:
‘Hmm…we really need those data in
by next month for that conference.’
PhD student Jim was summoned to his
supervisor Ella’s office to give a report
on his current progress. Jim asked for
an
extension
but
Ella
looked
concerned. Her face turned serious:
‘Hmm…we really need those data in
by next month for that conference.’
Indirect
speech
PhD student Ella was summoned to
her supervisor Jim’s office to give a
report on her current progress. Ella
asked for an extension but Jim looked
concerned. His face turned serious,
saying that they really needed those
data in by next month for that
conference.’
PhD student Jim was summoned to his
supervisor Ella’s office to give a report
on his current progress. Jim asked for
an
extension
but
Ella
looked
concerned. Her face turned serious,
saying that they really needed those
data in by next month for that
conference.’
Voice
focused
Voice
unfocused
Sixty octets of short fictitious stories containing different protagonists were prepared as
reading materials (The full stimulus set can be found in Appendix 2). Each story started
with two declarative sentences to set up a scenario, followed by a speech quotation from a
main protagonist in the story. The quotation employed either the direct speech or the
indirect speech reporting style (Reporting Style factor); each was preceded by either a
voice-focused description or a voice-unfocused description (Linguistic Focus factor). The
reported speaker‟s gender was either a male or a female speaker (Gender factor). The
reported speaker‟s emotional arousal in each item was rated by a different sample of
subjects (see 4.2.4.1) and the sixty items were evenly split into 20 „high‟, 20 „medium‟,
and 20 „low‟ emotional arousal groups (Emotion factor). The following are three examples
of the high, medium and low arousal stories:
[HIGH] Millionaire Joseph was addicted to betting on horses. Tipped by a so-called
„insider‟, he recently placed an enormous bet, but shockingly, the horse had lost. Angry with
his informant, Joseph shouted furiously on the phone: ‘Where did you bloody information
come from!? That was a huge amount of money – almost one million pounds!’
Page | 78
[MEDIUM] Britney is a student at the University of Glasgow. After a heavy snow in the
afternoon, she was complaining to her boyfriend James about the weather on their way home.
Her voice sounded very grumpy and unpleasant: ‘I really hate the winter! It’s always dark
and the roads are too slippery.’
[LOW] Smith was working in a small antiques shop down the local high street. Today, a
middle-aged posh lady with thick glasses came into the shop. She looked around and said in
a nonchalant tone: ‘You may be surprised to learn that I’m a world-renowned collector of
rare memorabilia of White-eared Pheasant.’
Eight lists of stimuli were created, each containing counterbalanced item-condition
combinations. Each item appeared once per list, but in a different condition across lists.
The presentation order of the items was randomised for each list.
4.2.3 Procedure
The experiment took place in a sound-attenuated recording room (Voice Neurocognition
Lab, CCNi). The participants were given one of the eight stimulus lists. They were
instructed to read the stories aloud as naturally and fluently as possible, and only for once;
such an instruction did not explicitly encourage participants to vocally „act out‟ the stories
during reading. After setting up the microphone and the digital recorder, the experimenter
sealed the room and the participant started to read the stories aloud while audio recordings
were taken using Audacity (http://audacity.sourceforge.net/). The recordings were saved
as .WAV files for later analysis.
4.2.4 Data analyses
4.2.4.1
Coding the reported speaker’s emotional arousal
A different sample of 9 native speakers of English (3 male) were recruited to rate the
reported speaker‟s emotional arousal for each story. They were presented with the whole
list of 60 stories.
The stories only contained direct speech quotations without the
Linguistic Focus manipulation (the voice-focused or –unfocused descriptions were all
replaced with He/She said; see the example below).
PhD student Ella was summoned to her supervisor Jim‟s office to give a report on her current
progress. Ella asked for an extension but Jim looked concerned. (Context)
He said: „Hmm…we really need those data in by next month for that conference.‟ (Directspeech)
Page | 79
The participants were instructed to read each story carefully and then rate, taking into
account the described situation, how excited/agitated they think the reported speaker would
be while he/she was speaking (the direct speech quotations were underscored in the actual
stimuli), using a 5-point scale: 0 meant „very calm or sad‟, 4 meant „very excited or
agitated‟. They wrote down the ratings on rating sheets; the latter were collected by the
experimenter for analysis.
The collected emotional arousal ratings were averaged by item and were then sorted in an
ascending order. The 20 items with the lowest mean ratings were allocated into the „low‟
emotional arousal group (M = 1.01, SD = 0.32) while the middle 20 items into the „medium‟
emotional arousal group (M = 1.95, SD = 0.34) and the final 20 items with the highest
average ratings into the „high‟ emotional arousal group (M = 3.05, SD = 0.43).
4.2.4.2
Acoustic analyses of
0
on the recordings of the critical quotations
Acoustic analyses of pitch (F0) were conducted in Praat software (Boersma & Weenink,
2010).
Advised by the Praat manual (http://www.fon.hum.uva.nl/praat/manual/), a
frequency range of 75 to 300 Hz was applied for male participants and a range of 100 to
500 Hz for female participants.
The audio samples of the critical direct and indirect speech quotations were listened to and
manually selected by the experimenter. The F0 values on the selected samples were
extracted by frame (i.e., by roughly every 7.5 ms), and were then averaged ( 0) for each
experimental trial. The
0
measures were not suitable for ANOVAs because they were
distributed bimodally due to the systematic pitch difference between the male participants
(M = 118 Hz) and the female participants (M = 205 Hz). To deal with this problem, they
were „centred‟ by subtracting their by-subject means such that the „centred‟ pitch means
(labelled as
0‟
0‟)
were normally distributed with a grand mean of 0 Hz. Next, trials with a
of more than 3 SD above a condition‟s mean were considered as outliers and were
removed. In total, 37 trials (1.3% of the trials, equally distributed across conditions) were
excluded. The
0‟s
in the remaining 2843 trials were submitted to 2 [Reporting style:
direct versus indirect speech] × 3 [Emotion (arousal): low versus medium versus high
emotional arousal of reported speakers] × 2 [Gender: male versus female reported speakers]
Page | 80
× 2 [Linguistic Focus: voice-focused versus voice-unfocused pre-speech descriptions]
ANOVAs with subject (F1) and item (F2) as random factors.
4.2.5 Results and discussion
There was a significant Reporting Style main effect, F1(1, 47) = 117.588, p < .001, F2(1, 59)
= 217.113, p < .001. It was found that participants significantly increased their pitch (F0)
during oral reading of direct speech (M = 4.71 Hz, SD = 12.63 Hz) as opposed to indirect
speech (M = -5.21 Hz, SD = 8.68 Hz). Although this finding was not central to our
hypothesis, it was nonetheless in line with the „inner voice‟ experience we had observed in
our previous studies; the latter showed that individuals are more likely to mentally simulate
vivid vocal depictions of reported speakers during comprehension of direct speech as
opposed to indirect speech (Chapter 2 & 3).
Likewise, mental simulations of vocal
depictions should also take place during oral reading (since it requires comprehension of
the written text) of direct speech rather than indirect speech.
Thus, the simulated
representations of vivid vocal depictions would then be orally interpreted in the form of
vivid voice acting of the reported speech acts via articulation. Such voice acting in oral
reading of direct speech was confirmed both by the experimenter‟s personal observation of
the recorded audio samples and by the significantly larger variation of pitch in the direct
speech condition than in the indirect speech condition (quantified in pitch standard
deviations over time, SD = 12.63 versus 8.68, paired-sample t(47) = 11.35, p < .001, t(59)
= 12.61, p < .001). Voice acting requires relatively stronger and more complex vocal
modulations of acoustics. To achieve this, oral readers have to increase their lung volume
so as well as the subglottal pressure to prepare extra vocal energy. The changes in the
subglottal pressure enhance the frequency of vocal fold vibration, thereby increasing the
overall F0. In this sense, the significant Reporting Style main effect could have been
engendered by engaging in voice acting during oral reading of direct speech rather than
indirect speech. Another plausible contributor to this main effect is the F0 resetting9 in oral
reading of direct speech but not indirect speech. In speech communication, F0 tends to
decline throughout the course of a declarative sentence (Bolinger, 1964; Cohen, Collier, &
Thart, 1982; Cohen & t' Hart, 1967). A single F0 declination could either span an entire
9
F0 resetting refers to the phenomenon where the declining F0 contour jumps up to a relatively higher F0
level such that the declination preceding and following the resetting can be better predicted by two separate
functions rather than one (see an example in Figure 9)
Page | 81
sentence (e.g., Figure 10) or be segmented into multiple „breath groups‟ (e.g., Figure 9).
Speakers replenish air between these „breath groups‟, each starting with a F0 resetting. In
communication, F0 declination and resetting encode important pragmatic information
(Danly et al., 1983; Hird & Kirsner, 2002; Menn & Boyce, 1982; Umeda, 1981; Vaissiere,
1983) such as the „units of meaning‟ which speakers plan to convey (which can span from
a single word to seven or more clauses (Winkworth, Davis, Adams, & Ellis, 1995). It was
observed that direct and indirect speech, as different „units of meaning‟ or „pragmatic
entities‟, were indeed characterised in differential breathing patterns during oral reading.
After reading the main clauses of direct and indirect speech (e.g., He said), oral readers
frequently paused before reading direct speech quotations but rarely do so before reading
indirect speech quotations. The pauses before reading direct speech quotations perhaps
signal a perspective switch from the reporter to the reported speaker. Nevertheless, such a
pause introduces a F0 resetting and increases the
0
for oral reading of the following direct
speech quotations (Figure 11). By contrast, without such a pause before oral reading of
indirect speech quotations, the F0 continues to decline from reading of the main clauses,
resulting in a relatively lower
0
for oral reading of the following indirect speech
quotations (Figure 11). Thus, the significant Reporting Style main effect may have also
been engendered by the differential breathing patterns during oral reading of direct speech
versus indirect speech. That said, since we were interested in whether the Reporting Style
factor interacted with other Context factors (i.e., Emotion and/or Gender), we did not
further validate or distinguish the above two interpretations of the Reporting Style main
effect.
Page | 82
Figure 9 Fundamental frequency (F0) contour (black) for sentence The cat that Sally owned was asleep
on the large branch in the tree. The blue line represents the prediction for a single F0 declination function
whereas the red lines represent the predictions for two declination functions. (Modified from Danly et al.,
1983)
Figure 10 Fundamental frequency (F0) contour (black) for sentence The book on the table was a gift
for my mother. The blue line represents the prediction for a single F0 declination function. (Modified from
Danly et al., 1983)
Page | 83
Figure 11 Hypothetical pitch (F0) contours (black) for direct speech Mary said: „My life has been amazing‟
(blue) and indirect speech Mary said that her life had been amazing (red). Due to the insertion of a F0
resetting, oral reading of direct speech elicits higher F0 on average than indirect speech.
In line with our predictions, we found that the Reporting Style factor significantly
interacted with the Emotion factor, F1(2, 46) = 40.478,.p < .001, F2(2, 58) = 11.789, p
< .001, and the Gender factor, F1(1, 47) = 6.236, p = .016, F2(1, 59) = 9.501, p = .003,
respectively. Exploring the former, we found that in the direct speech condition, there was
a consistent trend of increasing
0
The positive relationship between
as the reported speaker‟s emotional arousal increased.
0
and emotional arousal was consistent with the
acoustic profiles of vocal emotion expressions (Banse & Scherer, 1996). However, such
an increasing trend of
0
in relation to the reported speaker‟s arousal level was not
observed during oral reading of indirect speech (Figure 12). Exploring the Reporting Style
× Gender interaction, we found that oral readers significantly increased their pitch when
the reported speaker was female as opposed to when the reported speaker was male, but
only during oral reading of direct speech and not indirect speech (Table 3). The findings
indicate that individuals are more likely to adjust their own pitch in accordance with the
reported speaker‟s emotional arousal and gender during oral reading of direct as opposed
to indirect speech. It supports the observation that individuals try to simulate the pitch of a
reported speaker‟s voice during oral reading of direct speech as opposed to indirect speech.
That said, further evidence is needed to verify that the underlying representations of the
observed interactions are voice-related in nature.
Page | 84
Figure 12 Illustrations of the Reporting Style × Emotion interaction. The numbers indicate the condition
means (in the centred pitch 0‟). The error bars represent the 95% confidence intervals for the means. The
simple effects (both by subjects and by items) of Emotion are also shown (n.s. = not significant; * = p < .05;
** = p < .01; *** = p < .001).
Direct
speech
Female reported speaker
Male reported speaker
t(47)
By
subject
p
Gender
Contrast
t(59)
By
item
p
5.28 (14.00)
3.14 (13.15)
1.796
.079
3.209
.002
Reporting Style Contrast
By subject
By item
t(47)
p
t(59)
p
-5.04 (9.10) 10.971 .000 11.764 .000
-4.55 (9.05) 8.653 .000 10.823 .000
-1.054
.297 Note: Significant contrasts
-.833 are highlighted in bold.
.408
Indirect
speech
Table 3 The means of the centred pitch ( 0‟, in Hz, with standard deviations in parentheses) for the direct
versus indirect speech critical quotations across the Gender contexts. Also shown are results from 2-tailed
sample t-tests examining the simple effects of Reporting Style and Gender.
To this end, we examined the effects of Linguistic Focus on these two interactions. We
found that the voice-focused manipulation had no noticeable effect on the Reporting Style
× Emotion interaction; the three-way interaction was not significant, Fs < 1.6, ps > .2. The
results could mean that the underlying mental representations of the Reporting Style ×
Emotion interaction may not be voice-related in nature, e.g., it could have been engendered
by supra-modal representations of emotional arousal; however, it was also possible that the
Reporting Style × Emotion interaction was too robust to be affected by the Linguistic
Focus manipulation. Future studies would be needed to disentangle between these two
Page | 85
possibilities. By comparison, the voice-focused manipulation appeared to effectively mask
the Reporting Style × Gender interaction: the three-way interaction was significant by item,
F2(1, 59) = 4.103, p = .047, and was approaching significance by subject, F1(1, 47) = 2.719,
p = .106 10. We explored this three-way interaction by splitting our data by Linguistic
Focus – one data set contained quotations that were preceded with voice-focused
descriptions (e.g., Her voice turned serious) while the other contained quotations that were
preceded with voice-unfocused descriptions (e.g., Her face turned serious). Next, we
performed the same 2 Reporting Style × 2 Gender ANOVAs on these two data sets,
respectively. It was discovered that in the voice-unfocused data set, the Reporting Style ×
Gender interaction was significant, F1(1, 47) = 6.624, p = .013, F2(1, 59) = 12.914, p
= .001. Consistent with our main findings, it was found that oral readers significantly
increased their pitch when the reported speaker was female as opposed to when the
reported speaker was male, but only during oral reading of direct speech rather than
indirect speech11 (see the upper panel in Table 4). By contrast, in the voice-focused data
set, the Reporting Style × Gender interaction was not significant, Fs < .83, ps > .36; in both
the direct speech and indirect speech conditions, oral readers‟ pitch did not differ as a
function of the reported speaker‟s gender (see the lower panel in Table 4). The results
appear to be in line with the masking hypothesis (see 4.1 Introduction and Figure 8B).
They indicate that linguistic focus on voices can effectively „mask‟ the Gender-concordant
0
modulations during oral reading of direct speech as opposed to indirect speech whereas
the linguistic focus on non-voice aspects cannot. Such a „voice-selective‟ masking effect
suggests that the underlying mental representations of the Reporting Style × Gender
interaction are likely to be voice-related in nature, i.e., contain representations of pitch
information.
10
The three-way interaction was not significant by subject. This was perhaps due to the relatively large
variability within a relatively small sample size (48 subjects in relation to a 2 × 2 × 2 design result in mere 6
data points per item per condition), suggesting a lack of statistical power rather than a null effect.
11
The by-subject Gender contrast was marginally significant due to the relatively large variability within a
relatively small sample size (also see Footnote 10).
Page | 86
Voice Unfocused data set
Female speaker
(Her face turned serious)
Male speaker
(His face turned serious)
t(47)
By
subject
p
Gender
Contrast
t(59)
By
item
p
Voice Focused date set
Female speaker
(Her voice turned serious)
Male speaker
(His voice turned serious)
t(47)
By
subject
p
Gender
Contrast
t(59)
By
item
p
Indirect
speech
Reporting Style Contrast
By subject
By item
t(47)
p
t(59)
p
5.73 (14.16)
-5.87 (8.83)
9.005 .000
9.872 .000
2.85 (12.69)
-4.65 (8.34)
6.298 .000
8.545 .000
Direct
speech
1.963
.056
3.050
.003
Direct
speech
-1.744
.088 Note: Significant contrasts
-1.667 are highlighted in bold text
.101
Reporting Style Contrast
Indirect
By subject
By item
speech
t(47)
p
t(59)
p
5.88 (13.87)
-5.01 (9.33) 9.670
.000 10.286 .000
4.74 (13.55)
-5.29 (9.72) 8.618
.000 10.002 .000
.909
.368
1.300
.199
.214
.832 Note: Significant contrasts
.386 are highlighted in bold text
.701
Table 4 The means of the centred pitch ( 0‟, in Hz, with standard deviations in parentheses) for the direct
versus indirect speech critical quotations across the Gender contexts in the two data sets separated by
Linguistic Focus. Also shown are results from 2-tailed sample t-tests examining the simple effects of
Reporting Style and Gender in each data set.
In summary, we have demonstrated that oral readers spontaneously adjust their pitch in
accordance with the linguistically implied reported speaker‟s emotional arousal and gender
during reading aloud of direct speech but not (or less so) during reading aloud of indirect
speech.
The Gender-concordant pitch modulations during oral reading of direct as
opposed to indirect speech can be selectively masked by exerting linguistic focus on the
reported speaker‟s voices, indicating that the underlying representations are likely to
contain acoustic information of pitch. The results are in line with our previous findings
(Chapter 2 & 3), suggesting that individuals are more likely to mentally simulate voice
representations during comprehension of direct as opposed to indirect speech.
More
specifically, we show that pitch may be an acoustic aspect of such voice representations, at
least in relation to reported speakers‟ genders, and that mental simulations of voices are not
only reflected in brain-activation patterns, but also in modulations of articulation during
oral reading.
Page | 87
4.3
Experiments 4, 5 and 6
The following three experiments were carried out to investigate mental simulations of pitch
during silent reading of direct versus indirect speech, i.e., without the demand for
articulation. We did not investigate the effects of reported speakers‟ emotional arousal
because linguistic focus on voices did not have a noticeable effect on Emotion-concordant
pitch modulations during oral reading of direct versus indirect speech in Experiment 3.
This suggests that the Emotion effects on pitch modulations may be representationally
voice-unrelated. For this reason, we exclusively focused on the effects of reported speakers‟
genders on mental simulations of pitch representations in the following investigations; this
was based on the findings in Experiment 3 that the representations underlying the observed
Gender-concordant pitch modulations during oral reading of direct rather than indirect
speech are likely to be voice-related in nature since they could be selectively masked by
exerting linguistic focus on voices.
4.3.1 Participants
Twenty-four native speakers of English participated in each of the three experiments in
exchange for £5.
They had no reported reading or hearing disabilities.
A typical
experimental session took approximately 45 min.
4.3.2 Stimuli and design
A 2 [Reporting Style: direct versus indirect speech] × 2 [Gender Match: the reported
speaker‟s gender matches versus mismatches the probe word speaker‟s gender] design was
used in all three experiments. Sixty pairs of written stories with different protagonists
(indicated by different English names) were prepared as reading materials (see examples
below). They were almost the same as in Experiment 3, apart from that the Linguistic
Focus manipulation was removed and the Gender manipulation was now implemented
between items for the purpose of design simplification.
Page | 88
Gender Match
Gender Mismatch
Direct
speech
PhD student Ella was summoned to her
supervisor Jim’s office to give a report on
her current progress. Ella asked for an
extension but Jim looked concerned. He
said: ‘Hmm, we really need those data in
by next month for that conference.’
PROBE WORD: ‘conference’, male voice
PhD student Ella was summoned to her
supervisor Jim’s office to give a report on
her current progress. Ella asked for an
extension but Jim looked concerned. He
said: ‘Hmm, we really need those data in by
next month for that conference.’
PROBE WORD: ‘conference’, female voice
Indirect
speech
PhD student Ella was summoned to her
supervisor Jim’s office to give a report on
her current progress. Ella asked for an
extension but Jim looked concerned. He
said that they really needed those data in
by next month for that conference.
PROBE WORD: ‘conference’, male voice
PhD student Ella was summoned to her
supervisor Jim’s office to give a report on
her current progress. Ella asked for an
extension but Jim looked concerned. He
said that they really needed those data in by
next month for that conference.
PROBE WORD: ‘conference’, female voice
Each of the pair stories consisted of two declarative sentences to set up a scenario,
followed by either a direct speech or an indirect speech quotation (Reporting Style factor).
Half of the 60 items described utterances reported from female speakers while the other
half from male speakers. Spoken words were recorded as auditory probes for each item.
They were recorded from either a female speaker or a male speaker such that they either
matched or mismatched the reported speaker‟s gender indicated in the story (Gender
Match factor). To ensure that the male-speaking and the female-speaking spoken probe
words differed only in pitch rather than other acoustics, we normalised the intensity (dB)
of all probe word recordings using MatLab scripts (provided by the Voice Neurocognition
Laboratory, University of Glasgow), and adjusted the durations of the recording files (i.e.,
the lengths of the spoken words) and kept them proximate within each item. Acoustic
analyses on the edited probe words (see Table 5) confirmed that the male-speaking words
and the female-speaking words differed from each other only in pitch. While the malespeaking probe words were significantly lower in pitch than the female-speaking probe
words (Mdiff = 74.6 Hz), there were no meaningful/significant differences between them in
loudness or duration; although the male-speaking words were significantly louder than the
female-speaking words, the difference amounted to a mere 0.07 dB (in loudness perception
given that a 5 dB increment step size is usually recommended for subjects with normalhearing sensitivity, Cox, Alexander, Taylor, & Gray, 1997).
This indicates that our
intended manipulation of the pitch of the spoken words was effective. In addition to the
acoustic manipulations, we also controlled the presentation position of the corresponding
critical words in text to keep the presentations of the critical words and the spoken probe
words temporally close: We always selected, within a given speech quotation, the final
word as the critical word (e.g., the spoken word conference is the last word in He said:
Page | 89
„Hmm, we really need those data in by next month for that conference.‟); this was to ensure
that the spoken probe word would be presented soon enough after reading the critical word.
Also prepared were yes/no comprehension questions regarding the contents of the stories
(e.g., Was Ella Jim‟s PhD student?) which would assess participants‟ overall
comprehension performance and to ensure that they read the stories for comprehension.
Pitch (F0) (Hz)
Loudness (dB)
Duration (ms)
Male-spoken
probe words
Female-spoken
probe words
110.8 (13.4)
78.46 (0.35)
674 (119)
185.4 (17.9)
78.39 (0.25)
678 (118)
Paired-sample t-tests
by item
t(59)
p
-29.877
<.001
2.080
.042
1.628
.109
Table 5 Means for pitch, loudness and durations (along with standard deviations in parentheses) on
the audio recordings of the male-speaking and the female-speaking probe words. Also shown are
results for the paired-sample t-tests investigating the simple contrasts of the Probe Word Speaker‟s Gender.
In addition to the 60 critical items, 120 filler stories were prepared to conceal the intended
experimental manipulations.
The filler stories were mostly declarative narratives.
Twenty-eight of them (ca. 23%) contained direct speech (14 items) or indirect speech
sentences (14 items) but without the critical experimental manipulations. In Experiment 4
and 6, one third of the filler items (40 items) contained spoken probe words which matched
the critical words in the written stories. The remaining 80 filler items contained mismatch
spoken words which were synonyms to the relevant critical words. Together there were
100 word match items (60 critical items + 40 fillers) and 80 word mismatch items (fillers).
In Experiment 5 (lexical decision task), one third of the filler items (40 items) contained
English meaningful words which either matched or mismatched the critical words in the
written stories. The remaining 80 filler items contained non-words that were digitally
constructed using the recordings of the words. For example, the non-word “bendarving”
was created using the recordings of “banana” and “starving”. Overall, there were 100
English meaningful word items and 80 non-word items.
Four lists of counterbalanced item-condition combinations were generated. A given trial
could occur in one of the four conditions: direct speech – male probe word, indirect speech
– male probe word, direct speech – female probe word, and indirect speech – female probe
word. Each item appeared in one condition per list and in different conditions across lists.
Page | 90
Each participant only saw one list. The presentation order was randomised for each
participant.
4.3.3 Procedure
The experiment was conducted in a lab equipped with a Dual Core Dell PC. Participants
were seated in a comfortable chair, wearing headphones. Stimuli were presented using Eprime 2.0 presentation software (Psychology Software Tools, Inc., USA).
4.3.3.1
Procedure for Experiment 4
After a ten-trial practice, participants were presented with the given stimulus list in the
following manner. In each trial, participants first saw a cue „NEW TRIAL‟ at the centre of
the computer screen for 1000 ms, which was followed by the presentation of a short story.
The text was presented in a 15 pt Arial font on a light grey background. The whole text
passage was centred and wrapped within 50% width of the screen, together making the
reading stimuli appear as natural as possible. Participants were instructed to silently read
the text passage in their own time, and press spacebar to terminate the presentation when
they finished reading. Next, they saw a question mark at the centre of the screen for 1000
ms and then heard a spoken word through the headphones. Participants were required to
indicate whether the spoken word had been mentioned in the story they had just read as
quickly and accurately as possible by pressing the left mouse button for „yes‟ or the right
one for „no‟. Their reaction times (RTs) and recognition accuracy on the spoken words
and were recorded. Finally, in one third of the trials, they were required to answer a yes/no
comprehension question regarding the story they had just read. The answering accuracy
was also recorded.
4.3.3.2
Procedure for Experiment 5
The procedure was the same as Experiment 4 except for that in Experiment 5, a lexical
decision task was employed: As participants heard a spoken probe word via the
headphones, they had to indicate, as quickly and accurately as possible, whether the heard
spoken word was an English meaningful word or not, by pressing the left mouse button for
„yes‟ or the right mouse button for „no‟.
Page | 91
4.3.3.3
Procedure for Experiment 6
Instead of using self-paced presentation of text for silent reading, Experiment 6 employed
serial presentation of text. The reading materials were segmented and presented in serial.
For example, direct speech sentence He said: „Hmm, we really need those data in by next
month for that conference.‟ was segmented (the boundaries are indicated by „//‟) into He
said: „//Hmm,//we really need// those data in// by next month// for// that// conference.‟ The
presentation duration of each segment was determined by 100 ms + 50 ms × number of
syllables + 200 ms × number of words, allowing sufficient time for reading. Critically, the
presentation would pause on the critical word (e.g., the final word „conference‟) while a
spoken probe word would be presented simultaneously through the headphones.
Participants had to respond as quickly and accurately as possible, using the mouse button
(left for „yes‟, right for „no‟), as to whether the word that they had read and the word that
they had heard were the same or not.
4.3.4 Results and discussion
Incorrect responses to the spoken words were discarded (2.5% of the trials in Experiment 4,
0.7% and 2.3% in Experiment 5 and 6, respectively). Outliers were excluded in two steps:
First, the correct responses with a RT of longer than 2000 ms or shorter than 250 ms were
removed. Next, the remaining trials with a RT which was 3SD away from a condition‟s
mean were excluded.
In total, 5.5% of the trials were excluded from analysis in
Experiment 4, 2.5% and 4.7% in Experiment 5 and 6, respectively.
The remaining trials were submitted to 2 [Reporting Style: direct versus indirect] × 2
[Gender Match: the probe word‟s gender matched versus mismatched the reported
speaker‟s gender] ANOVAs with subject (F1) and item (F2) as random factors, respectively.
The accuracy on comprehension questions was above 70% across all three experiments
(Experiment 4: M = 86.7%, SD = 6.2%; Experiment 5: M = 87%, SD = 7.8%; Experiment
6: M = 86.8%, SD = 7.6%), indicating that subjects comprehended the reading materials
attentively to a reasonable extent.
We did not find any significant effect or interaction in any of three experiments, except
that in Experiment 6 (using the immediate probing paradigm where the written critical
Page | 92
word and the spoken probe word were presented simultaneously), the Reporting Style ×
Gender Match interaction was approaching significance, F1(1, 23) = 3.369, p = .079, F2(1,
59) = 2.454, p = .123. Exploring this interaction, it was found that in the direct speech
condition, participants‟ response times (RTs) to the spoken probe words were faster when
the probe word speaker‟s gender matched the reported speaker‟s gender (M = 742 ms) as
opposed to when they mismatched (M = 760 ms). Such a match advantage in RTs is in
line with mental simulations of pitch representations during silent reading of direct speech:
The recognition of the gender-match spoken words was facilitated as compared to the
recognition of the gender-mismatch spoken words because the former contained pitch
information that was more congruent with the mentally simulated pitch representations
during silent reading. By contrast, in the indirect speech condition, the RT difference
across the Gender Match manipulation in the indirect speech condition was in the opposite
direction (744 ms versus 736 ms for the match versus mismatch conditions). This match
disadvantage was inconsistent with the pitch simulation hypothesis and amounted to only
half as much as compared to the direct speech condition (the RT difference = 8 ms in
indirect speech versus 18 ms in direct speech). Such result patterns seemed to resemble a
gender-match advantage in silent reading of direct speech and a null effect in silent reading
of indirect speech, which was in line with our prediction. Thus, we suspected that there
may have been latent factors which may have diminished the significance of the Reporting
Style × Gender Match interaction.
Indeed, in post-hoc analyses of Experiment 6 (immediate probing), we found that the
Reporting Style × Gender Match interaction seemed to interact with the Reported
Speaker‟s Gender (the three-way interaction was approaching significance, F1(1, 23) =
3.347, p = .080, F2(1, 59) = 1.863, p = .178). We separated the data into two groups: One
group consisted of items with female reported speakers while the other with male reported
speakers. We then performed the same Reporting Style × Gender Match ANOVAs on the
two sets of data, respectively. We found that when the reported speaker was female, the
Reporting Style × Gender Match interaction was significant both by subject and by item,
F1(1, 23) = 4.890, p = .037, F2(1, 59) = 5.036, p = .032, whereas when the reported speaker
was male, this interaction was not significant, Fs < .14, ps > .7. Exploring the former, we
found that in silent reading of direct speech, participants‟ RTs to the spoken probe words
were significantly faster when the probe word speaker‟s gender matched the reported
Page | 93
speaker‟s gender than when they mismatched; whereas in silent reading of indirect speech,
such Gender Match effects were not observed (Table 6).
Female-speaking items
Gender Match
Gender Mismatch
t(23)
By
Gender
subject
p
Match
t(29)
By
Contrast
item
p
Male-speaking items
Gender Match
Gender Mismatch
t(23)
By
Gender
subject
p
Match
t(29)
By
Contrast
item
p
Reporting Style Contrast
By subject
By item
t(23)
p
t(29)
p
-1.381 .181 -.800 .430
1.718 .099 2.036 .051
Direct
speech
Indirect
speech
741 (192)
782 (242)
-2.524
.019
-2.030
.05
759 (194)
745 (217)
.826
.418 Note: significant contrasts
.630 are highlighted in bold text
.534
Reporting Style Contrast
Indirect
By subject
By item
speech
t(23)
p
t(29)
p
732 (188)
.402 .692 .317 .753
724 (203)
.695 .494 .595 .556
.937
.358 Note: significant contrasts
.172 are highlighted in bold text
.865
Direct
speech
741 (180)
738 (204)
.237
.815
-.022
.983
Table 6 Reaction times (in ms, with standard deviations in parentheses) for the direct versus indirect
speech conditions across the Gender Match manipulations in the two data sets separated by the gender
of the reported speaker (Experiment 6). Also shown are results from 2-tailed sample t-tests examining the
simple effects of Reporting Style and Gender Match in each data set.
The results show that individuals are more likely to mentally simulate voice
representations (which contain pitch information) in accordance with the reported
speaker‟s gender during silent reading of direct as opposed to indirect speech, but only
when these utterances are reported from a female speaker. Such „female-favoured‟ voice
simulations could perhaps be explained by women‟s superior expressivity over men during
verbal communication (e.g., Kring & Gordon, 1998; Robinson & Johnson, 1997): Female
speakers are more likely to employ vivid vocal depictions in verbal communication than
their male counterparts. Thus, comprehension of female-speaking direct speech should be
grounded in mental simulations of more vivid vocal depictions that are employed by
female speakers whereas comprehension of male-speaking direct speech would be
grounded in experiences of not-so-vivid utterances by male speakers. This could explain
why individuals are more likely to simulate voices when reading female-speaking direct
speech as opposed to male-speaking direct speech. Although the three-way (Reported
Speaker‟s × Gender Reporting Style × Gender Match) interaction did not reach .05
Page | 94
significance, the post-hoc analyses nonetheless provide a potential hint as to how
experiences with differential expressivities between male and female speakers underlie
language comprehension. To further validate this „gender-dependent‟ voice simulation
phenomenon, future research will need to replicate the current results with more strictly
counterbalanced design and more carefully manipulated stimuli. For example, the reported
speaker‟s gender may be manipulated within each item to rule out the potential itemspecific factors that may have been confounded with the reported speaker‟s gender factor
(for design simplification, the current investigations merely manipulated the reported
speakers‟ gender between items).
It should be noted that this gender-dependent voice simulation phenomenon in Experiment
6 (immediate probing) was the only effect that we observed in all three experiments. It
may reflect that the simulated voice representations during silent reading of direct rather
than indirect speech are „short-lived‟, similar to mental simulations of actions (Kaschak &
Borreggine, 2008).
This could perhaps explain why there was no effect detected in
Experiment 4 or 5: The „off-line‟ paradigms we employed in Experiments 4 and 5 present
spoken probe words 1000 ms after the offset of silent reading; this setup allows „shortlived‟ voice representations to „decay‟ before the presentation of the spoken probe words,
and it is therefore not sensitive enough to detect such representations. By contrast, the „online‟ paradigm we employed in Experiment 6 presents the written critical word and the
spoken probe words simultaneously; this setup does not allow the target representations to
„decay‟, and therefore it was sensitive enough to detect the gender-dependent voice
simulations. Although this higher detecting sensitivity of „on-line‟ paradigms was no more
than speculations, we nevertheless believed it would be sensible to exclusively focus on
„on-line‟ paradigms for our following investigations (e.g., Chapter 5 & 6).
In sum, Experiments 4 - 6 investigated mental simulations of pitch information during
silent reading of direct versus indirect speech. We examined the effects of the reported
speaker‟s gender on silent readers‟ recognition latencies to the gender-congruent and
gender-incongruent spoken words. Using an immediate probing paradigm in Experiment 6,
we found that when the reported speakers are female (e.g., She said) participants responded
faster to spoken words that are spoken by a same-gender speaker (i.e., a female speaker)
than to spoken words that are spoken by an opposite-gender speaker (i.e., a male speaker)
during silent reading of direct speech but not indirect speech; when the reported speakers
Page | 95
are male (e.g., He said), such Gender Match effects were not observed. The results
suggest that individuals are more likely to mentally simulate pitch information in
accordance with the reported speaker‟s gender during silent reading of direct as opposed to
indirect speech, but only when the reported speaker was female.
4.4
Summary and general discussion
The current investigation made the first attempt at decomposing the „inner voice‟
experiences that we had observed during comprehension of direct versus indirect speech
into acoustic parameters. Specifically, we examined the pitch aspect of such „inner voice‟
representations during both oral reading (Experiment 3) and silent reading (Experiments 46) of direct versus indirect speech.
Experiment 3 showed that individuals spontaneously adjusted their own pitch in
accordance with the reported speaker‟s emotional arousal and gender (as derived from the
context in which the reported utterances were embedded) during oral reading of direct as
opposed to indirect speech. The gender-concordant pitch modulations can be masked by
exerting linguistic focus on reported speaker‟s voices, indicating that the underlying mental
representations are voice-related in nature and contain pitch information. The findings are
in line with our prediction, suggesting that individuals may mentally simulate the reported
speaker‟s pitch representations during oral reading of direct speech but not indirect speech.
Experiments 4 – 6 investigated mental simulations of pitch representations in silent reading
of direct versus indirect speech. Participants silently read story-embedded direct and
indirect speech and responded to male-speaking or female-speaking spoken words which
either matched or mismatched the reported speaker‟s gender described in the stories. It
was found that participants responded faster to the gender-match spoken words than to the
gender-mismatch spoken words during silent reading of direct speech rather than indirect
speech, but only when the reported speaker was female.
The results, although still
preliminary, suggest that the acquired experience with the differential expressivities
between direct and indirect speech acts as well as the ones between male and female
speakers underlies language comprehension.
Page | 96
A previous study on auditory imagery (Abramson, 2007) observed similar genderconcordant „inner voice‟ experiences during silent reading. In this study, participants were
first familiarised with a male voice and a female voice. They were then instructed to
silently read sentences, and had their implicit (auditory lexical decision) or explicit
memory (auditory recognition) for words assessed afterwards.
In the silently read
sentences, the gender of speaker was implied in the initial words, e.g., „He said,...‟ or „She
said,...‟.
Auditory lexical decision priming was found for gender-congruent items
following silent reading, but only up to 5 min after silent reading. In a second study,
similar lexical decision priming was found following listening to sentences, although these
effects remained reliable after a 2-day delay. These results suggest that readers create
episodic traces of text from auditory images of silently read sentences as they do during
listening.
The gender-congruency effects observed in this study also indicate the
possibility that pitch representations (given its critical role in signalling gender information)
might be encoded in silent reading.
That said, although similar, the „inner voice‟ phenomenon observed in Abramson‟s study
and the „inner voice‟ phenomenon observed in our investigations may reflect distinct
cognitive processes. The former may be interpreted as a comprehension-independent,
encouraged auditory imagery process whereas the latter is a spontaneous auditory
simulation process which is an integral part of language comprehension. In Abramson‟s
study, participants were first familiarised with two voices of different genders before the
actual reading trials, followed by cues (e.g.., He said) to the identity of the to-be-imagined
speaker within the actual reading materials. This type of experimental manipulation (i.e.,
pre-exposure to specific voices and/or imagery cues during reading) arguably encourages
participants to imagine speaker-specific voices during silent reading.
Hence, this
experimental setup was limited in determining whether such an auditory imagery would
also occur as a result of „normal‟ language comprehension.
By contrast, in our
investigations, participants were not presented with any specific voice samples before
reading. During reading, they read stories which contained only fictitious protagonists that
they did not know. Under this experimental setup, participants were unlikely to imagine
any specific, familiar voice. Nevertheless, they still activated voice-like representations
during reading, which may be better interpreted in mental simulations or a more
spontaneous imagery of unspecific, partial voice representations, as an integral part of
Page | 97
language comprehension. Moreover, our investigations also demonstrate that such an
„inner voice‟ experience can not only be modulated by linguistic cues to the reported
speaker‟s gender, but can also be modulated by pragmatically distinct reporting styles (i.e.,
direct versus indirect speech) and linguistic focus (on voices versus on non-voice aspects
of the reported speech acts). Thus, our findings may be better interpreted, in line with the
demonstration theory of direct speech (Clark & Gerrig, 1990) and the embodied theories of
language comprehension (Barsalou, 1999), as mental simulations of gender-concordant
pitch information that are guided by language comprehension, rather than as auditory
imagery of voices that are encouraged by non-linguistic factors.
In conclusion, our investigations address the pragmatics of different quotation styles and
their associated mental representations during text comprehension.
Our experiments
showed that oral readers spontaneously adjust their pitch to contextually implied pitch
when reading direct speech as opposed to meaning-equivalent indirect speech quotations,
and that silent readers respond differentially to spoken words that are spoken by samegender (as the reported speaker) or different-gender speakers but only when reading
female-speaking direct speech quotations.
These behavioural phenomena (although
requiring further validation) may be interpreted as perceptual simulations of reported
speakers‟ voices during comprehension of direct speech rather than indirect speech.
Specifically, it suggests that frequency representations such as pitch might be an acoustic
aspect of such simulated „inner voice‟ experiences during comprehension of direct speech
as opposed to indirect speech.
Page | 98
Chapter 5 Contextual modulations of loudness representations in
oral and silent reading of direct versus indirect speech
5.1
Introduction
In parallel with our investigations on the pitch aspect in voice simulations, the
investigations of the current chapter examined another acoustic aspect in such voice
simulation processes, namely the contextually implied loudness of the quoted speech.
Similar to pitch, loudness is an important acoustic characteristic of a speaker‟s voice for it
signals various types of information about a speaker (e.g., the speaker‟s emotional state;
Banse & Scherer, 1996) and his/her relations to the surrounding environments (e.g., the
surrounding noise level and the physical distance between the speaker and the listener).
We reasoned that if the simulated „inner voice‟ entails acoustic information such as pitch,
similar to the voices in the physical world, it is likely that such an „inner voice‟ may also
encode loudness representations. Thus, we employed the paradigms that we had developed
in Chapter 4 and investigated contextually-mediated mental simulations of loudness in
both oral reading (Experiment 7) and silent reading (Experiment 8) of direct versus indirect
speech.
Experiment 7 investigated whether individuals would adjust their own loudness (measured
in mean intensity (dB)) in accordance with the contextually implied loudness of the
reported speaker during oral reading of direct as opposed to indirect speech. Participants
orally read a series of written stories containing either direct or indirect speech while audio
recordings were taken for acoustic analyses.
It was predicted that they would
spontaneously increase their reading loudness when the preceding linguistic context
implied a loud-speaking reported speaker as opposed to a quiet-speaking reported speaker
during oral reading of direct speech but not indirect speech. Unlike Experiment 3 which
manipulated the reported speaker‟s pitch level in relation to his/her emotional arousal and
gender specifically, Experiment 7 manipulated the reported speaker‟s loudness level by
combining various types of linguistic information including reference to the environmental
noise level (e.g., party music), loudness-related verbs (e.g., shouted) and adverbs (e.g.,
loudly), and the reported speaker‟s emotional state (e.g., angry). We were not interested in
the individual effects of each type of linguistic information on reported speakers‟ loudness;
instead, we were only interested in the joint effects of these linguistic manipulations of the
Page | 99
reported speaker‟s loudness level. Thus, for the purpose of design simplification, we
collapsed across these linguistic manipulations into a single „loudness‟ manipulation; we
also dropped Experiment 3‟s Linguistic Focus manipulation12.
Experiment 8 employed the silent reading paradigm from Experiment 6 (i.e., the
immediate probing paradigm in which the written critical word and the spoken probe word
are presented simultaneously).
We examined the effects of the implied loudness of
reported speaker on the recognition latencies to a spoken probe word when the word‟s
loudness was made relevant by the previous text passage. Participants silently read short
stories containing direct or indirect speech utterances. The story was segmented first and
then presented in a serial fashion. The presentation of direct or indirect speech quotations
paused on a critical word while a spoken word was presented simultaneously. Participants
had to answer „yes‟ or „no‟ to indicate whether the spoken word they heard was the same
word as the one they saw on the screen. The spoken probe words were acoustically
manipulated into either a loud or a quiet version such that it either matched or mismatched
the loudness of the reported speaker that was linguistically implied in the story. We
acknowledge that the loudness of a speaker‟s voice may not be solely determined by the
absolute volume of a voice. For instance, one can shout quietly or shout from a distance.
In these situations, although the voices are low in volume, they would still be perceived as
loud. However, it would be impractical to consider possible factors which determine vocal
loudness all together in one go. For the purpose of simplification, we only focused on the
effects of vocal volumes (still labelled as loudness) and manipulated our spoken probe
words along this dimension in our current investigation. We predicted that participants‟
recognition on loudness-congruent spoken words (i.e., the reported speaker‟s loudness is
congruent with the spoken word‟s loudness) should be facilitated as opposed to recognition
on loudness-incongruent spoken words or vice versa.
12
The exclusion of the Linguistic Focus manipulation reduced the required subject number by 50%, which
significantly helped facilitate the progress of this research, especially under the tight time schedule.
Although we were no longer able to address the acoustic nature of the mental representations that underlie
mental loudness simulations in the current investigation, we could still address this question in the silent
reading study and in the future fMRI studies. Thus, for the sake of getting preliminary results as quickly as
possible, this exclusion was the optimal choice.
Page | 100
5.2
Experiment 7
5.2.1 Participants
Twelve native speakers of English (6 male) with no reported reading impairments
participated in exchange for £2. A typical session lasted about 15 minutes.
5.2.2 Stimuli and design
A 2 [Reporting Style: direct versus indirect speech] × 2 [Context: loud-speaking versus
quiet-speaking reported speaker] design was used in this experiment. Twenty-four pairs of
short fictitious stories with different protagonists were prepared as reading materials (see
below for an example and see Appendix 3 for the full stimulus set). Each story contained
either a direct speech or an indirect speech quotation (Reporting Style factor). As for each
pair, one story described a situation where the protagonist would speak loudly whereas the
other a situation where the speaker would speak more quietly (Context factor). Crucially,
the critical sentence (bold texts in the following example) within each reporting style was
identical across contexts and virtually the same across reporting styles.
Loud-speaking context
Quiet-speaking context
Direct
speech
Katy’s friends were having such a great time
blasting cheesy Christmas music in her flat
that she barely heard the knock at the door at
1am. She was expecting another friend and
was surprised to find a policeman looking
crossly over her shoulder at the party going on
inside. He looked very serious, demanding:
‘Excuse me Madam. I suggest you keep the
noise down as there's been a complaint
from your neighbours.’
Katy was reading in the library and had put ear
plugs in to block out the other students studying
nearby. She absentmindedly began to click her
tongue, not realising that other people could
hear her until a concerned librarian tapped her
lightly on the shoulder. He politely said: ‘Excuse
me Madam. I suggest you keep the noise
down as there's been a complaint from your
neighbours.’
Indirect
speech
Katy’s friends were having such a great time
blasting cheesy Christmas music in her flat
that she barely heard the knock at the door at
1am. She was expecting another friend and
was surprised to find a policeman looking
crossly over her shoulder at the party going on
inside. He looked very serious, demanding
that she should keep the noise down as
there had been a complaint from her
neighbours.
Katy was reading in the library and had put ear
plugs in to block out the other students studying
nearby. She absentmindedly began to click her
tongue, not realising that other people could
hear her until a concerned librarian tapped her
lightly on the shoulder. He politely said to her
that she should keep the noise down as
there had been a complaint from her
neighbours.
Four stimulus lists with counterbalanced item-condition combinations (6 stories per
condition per list) were generated. Each item appeared once per list, but in a different
condition across lists. The lists were word-processed into four written scripts. The order
Page | 101
of the stories per script was randomised. Each script was randomly assigned to a quarter of
our participants.
5.2.3 Procedure
The experiment was conducted in a sound-attenuated room. Participants were instructed to
read out the stories from the script in one go and as naturally and fluently as possible. The
instructions did not explicitly encourage participants to vocally „act out‟ the stories during
reading.
After setting up the microphone and the digital voice recorder, the experimenter sealed the
room and the participant started reading the stories aloud while audio recordings were
taken using Audacity (http://audacity.sourceforge.net/).
The recordings were saved
as .WAV files for later analyses.
5.2.4 Results and discussion
Acoustic analyses of intensity (dB) were conducted in Praat (Boersma & Weenink, 2010).
The audio recordings of the critical sentences were listened to and manually selected by the
experimenter. The intensity values on these recordings were extracted by frame, i.e., by
roughly every 7.5 ms. In line with the standards for intensity analysis (according to Praat
manual; http://www.fon.hum.uva.nl/praat/manual/), only the intensity values between 50
dB and 100 dB were considered as valid for the analysis. They were first averaged by
trials and the by-trial intensity means were then submitted to 2 [Reporting Style: direct
versus indirect speech] × 2 [Context: loud-speaking versus quiet-speaking reported speaker]
ANOVAs with subject (F1) and item (F2) as random factors.
The Reporting Style × Context interaction was significant by subject, F1 (1, 11) = 11.85, p
= 0.006, and marginally significant by item, F2 (1, 23) = 3.805, p = 0.06313. Exploring the
interaction, we found that direct speech quotations were read out significantly more loudly
13
The marginally significant interaction by item, as compared with the significant by-subject interaction,
suggests that there might have been „nuisance‟ items in which the loudness manipulations were not as
effective as intended (presumably due to our more unspecific ways of manipulating loudness). Nevertheless,
given that the interaction pattern was consistent with our hypothesis and the critical pairwise contrasts were
significant as expected, this marginally significant interaction by item may simply reflect the lack of
statistical power rather than a null effect.
Page | 102
when the preceding linguistic context implied a loud-speaking as opposed to a quietspeaking reported speaker. By contrast, such contextual modulations of loudness were not
found in oral reading of indirect speech (Table 7). The results showed that our participants
adjusted their loudness in accordance with the linguistically implied loudness of the
reported speaker during oral reading of direct speech as opposed to indirect speech. The
results cannot be explained by differences in phonological or morphological features (tense,
person, etc.) between the two Reporting Style conditions because the interaction involved
differential effects of Context within each reporting style (the critical sentences were
identical). The current findings are in line with our prediction, suggesting that oral readers
may mentally simulate the reported speaker‟s voice representations during oral reading of
direct as opposed to indirect speech. More specifically, we show that loudness may be
another acoustic aspect of such voice representations, and that mental simulations of voice
are not only reflected in brain-activation patterns, but also in modulations of vocal
acoustics during oral reading. However, without the Linguistic Focus manipulations (cf.
4.1 Introduction for the rationale), we did not have the relevant information to verify the
voice-related nature of the underlying mental representations. Therefore, we would have
to address this question in silent reading (Experiment 8) and in our future investigations.
Direct
Speech
Loudly-speaking Context
Quietly-speaking Context
t(11)
By
subjects
p
Context
Contrast
t(23)
By
items
p
Indirect
Speech
Reporting Style Contrast
By subjects
By items
t(11)
p
t(23)
p
4.423
.001
5.147 < .001
.858
.4
2.722
.02
66.3 (4.5) 63.7 (3.4)
63.7 (3.9) 63.2 (3.8)
1.497
3.632
.163 Note: Significant contrasts are
.004
.631 highlighted in bold text
3.306
.534
.003
Table 7 The intensity means (in dB, along with standard deviations in parentheses) across Context on
the critical direct versus indirect speech quotations in Experiment 7. Also shown are results from 2tailed paired sample t-tests examining the simple effects of Context and Reporting Style.
Page | 103
5.3
Experiment 8
5.3.1 Participants
Sixty-four native speakers of English with no reported reading or hearing impairments
participated in exchange for £3.
A typical session took around 20 minutes.
Three
participants had to be excluded from analysis because their accuracy on the comprehension
questions was below 70%, the latter suggesting that they did not read the story attentively.
The remaining sixty-one subjects (41 female, 20 male, age 17 – 37) were valid for the
following analyses.
5.3.2 Stimuli and design
A 2 [Reporting Style: direct versus indirect speech] × 2 [Context: loud-speaking versus
quiet-speaking reported speaker] × 2 [Word: loud-speaking versus quiet-speaking spoken
word] design was used in this experiment (see the example below; next page).
Twenty-four pairs of short fictitious stories with different protagonists were prepared as
reading materials (they were exactly the same as those in Experiment 7). Each story
contained either a direct speech or an indirect speech quotation (Reporting Style factor).
As for each pair, one story described a situation where the protagonist would speak loudly
whereas the other a situation where the speaker would speak quietly (Context factor).
Crucially, the critical sentence (bold texts in the following example) within each quotation
was identical across Contexts and virtually the same across Reporting Styles. Recordings
of spoken words were taken from a female native English speaker and prepared as the
probe stimuli. As for the 24 critical items, the spoken words matched the final words in
the speech quotations (e.g., the spoken word „neighbours‟ matches the last word in He
looked very serious, demanding: „Excuse me Madam. … from your neighbours.‟). They
were acoustically normalised and manipulated into „loud‟ and „quiet‟ versions such that
they either matched or mismatched the reported speakers‟ loudness which was contextually
described in the stories (Word factor). Since the two versions of each spoken word were
derived from the same recording file, they matched in all acoustic characteristics and only
differed in loudness (i.e., vocal volume in the current investigation). The latter was
confirmed by acoustic analysis on intensity: The loud-speaking words (M = 73.1 dB, SD =
Page | 104
9.6) were significantly louder than the quiet-speaking words (M = 56.6 dB, SD = 9.6),
paired-sample t(23) = 870.9, p < .001.
Direct
speech
Loudspeaking
context
Indirect
speech
Direct
speech
Quietspeaking
context
Indirect
speech
Loud-speaking word
Quiet-speaking word
Katy’s friends were having such a great
time blasting cheesy Christmas music in
her flat that she barely heard the knock
at the door at 1am. She was expecting
another friend and was surprised to find
a policeman looking crossly over her
shoulder at the party going on inside.
He looked very serious, demanding:
‘Excuse me Madam. I suggest you
keep the noise down as there's been
a complaint from your neighbours.’
Katy’s friends were having such a great
time blasting cheesy Christmas music in
her flat that she barely heard the knock
at the door at 1am. She was expecting
another friend and was surprised to find
a policeman looking crossly over her
shoulder at the party going on inside.
He looked very serious, demanding:
‘Excuse me Madam. I suggest you
keep the noise down as there's been
a complaint from your neighbours.’
Loud spoken probe word
Quiet spoken probe word
(e.g., ‘neighbours’)
Katy’s friends were having such a great
time blasting cheesy Christmas music in
her flat that she barely heard the knock
at the door at 1am. She was expecting
another friend and was surprised to find
a policeman looking crossly over her
shoulder at the party going on inside.
He looked very serious, demanding that
she should keep the noise down as
there had been a complaint from her
neighbours.
(e.g., ‘neighbours’)
Katy’s friends were having such a great
time blasting cheesy Christmas music in
her flat that she barely heard the knock
at the door at 1am. She was expecting
another friend and was surprised to find
a policeman looking crossly over her
shoulder at the party going on inside.
He looked very serious, demanding that
she should keep the noise down as
there had been a complaint from her
neighbours.
Loud spoken probe word
Quiet spoken probe word
(e.g., ‘neighbours’)
Katy was reading in the library and had
put ear plugs in to block out the other
students
studying
nearby.
She
absentmindedly began to click her
tongue, not realising that other people
could hear her until a concerned
librarian tapped her lightly on the
shoulder. He politely said: ‘Excuse me
Madam. I suggest you keep the noise
down as there's been a complaint
from your neighbours.’
(e.g., ‘neighbours’)
Katy was reading in the library and had
put ear plugs in to block out the other
students
studying
nearby.
She
absentmindedly began to click her
tongue, not realising that other people
could hear her until a concerned
librarian tapped her lightly on the
shoulder. He politely said: ‘Excuse me
Madam. I suggest you keep the noise
down as there's been a complaint
from your neighbours.’
Loud spoken probe word
Quiet spoken probe word
(e.g., ‘neighbours’)
Katy was reading in the library and had
put ear plugs in to block out the other
students
studying
nearby.
She
absentmindedly began to click her
tongue, not realising that other people
could hear her until a concerned
librarian tapped her lightly on the
shoulder. He politely said to her that
she should keep the noise down as
there had been a complaint from her
neighbours.
(e.g., ‘neighbours’)
Katy was reading in the library and had
put ear plugs in to block out the other
students
studying
nearby.
She
absentmindedly began to click her
tongue, not realising that other people
could hear her until a concerned
librarian tapped her lightly on the
shoulder. He politely said to her that
she should keep the noise down as
there had been a complaint from her
neighbours.
Loud spoken probe word
Quiet spoken probe word
(e.g., ‘neighbours’)
(e.g., ‘neighbours’)
Page | 105
Moreover, 36 filler stories containing no direct or indirect speech were prepared and mixed
with the 24 critical items to conceal the experimental manipulations of the study. The
spoken words for the 36 filler items either matched or mismatched the selected critical
words in the stories (not always the last words) and they were also presented in either a
„loud‟ or a „quiet‟ version. Content-related comprehension questions were made for all
items to encourage reading for comprehension.
All reading materials except
comprehension questions were segmented for serial visual presentation.
Eight lists of stimuli were generated with counterbalanced item-condition combinations
(12 direct speech and 12 indirect speech items per list; 6 loudness match (3 loud, 3 quiet)
and 6 loudness mismatch items (3 loud, 3 quiet) in each reporting style condition). Each
critical item appeared only once per list, but in a different condition across lists. The eight
lists of stimuli were evenly allocated to the participants. The presentation order of the
stories on each list was randomised for each participant.
5.3.3 Procedure
The procedure was the same as in Experiment 6. Participants were seated comfortably in
front of a Dell Dual Core PC, wearing headphones for spoken probe word presentation.
Each participant was assigned to one of the eight lists containing different versions of
experimental items, randomly interspersed with the filler stories. Stimulus presentation
was implemented in E-prime v2.0 software (Psychology Software Tools, Inc., USA).
Reading materials were presented in an 18 pt „Arial Narrow‟ font printed in black on a
light grey background. Each segment of text was presented at the centre of the screen.
Each trial began with the presentation of „NEW TRIAL‟ in the middle of the screen for
1000 ms. It was followed by serial visual presentation of segmented text passages for
silent reading. The presentation duration of each text segment was determined by 100 ms
+ 50 ms × the total number of syllables + 200 ms × the total number of words; this allowed
sufficient time for reading. The text presentation would pause on a critical word while a
spoken probe word was simultaneously presented. Participants had to respond as quickly
and accurately as possible, using the mouse buttons (the left mouse button for „yes‟, the
right for „no‟), as to whether the word that they had read and the word that they had heard
were the same or not. Their response times and accuracy were recorded. After reading,
Page | 106
participants had to answer a yes/no comprehension question regarding the content of the
story they had read via pressing the corresponding mouse buttons. Answering the question
triggered the presentation of the next trial.
Before the actual experiment, each participant had a 10-trial practice.
In the actual
experiment, they were allowed to have a short break after 30 trials.
5.3.4 Results and discussion
Incorrect responses to the spoken words were discarded (2.1%).
Outliers were then
excluded in two steps. First, the correct responses with a RT of longer than 2000 ms or
shorter than 250 ms were removed. Next, the remaining trials with a RT which was 3SD
away from a condition‟s mean were excluded. In total, 49 (3.3%) of the 1464 trials were
excluded from analysis. The remaining 1415 trials were submitted to 2 [Reporting Style:
direct versus indirect speech] × 2 [Context: loud-speaking versus quiet-speaking reported
speaker] × 2 [Word: loud versus quiet versions of the probe words] ANOVAs with subject
(F1) and item (F2) as random factors, respectively.
We found that the critical three-way interaction Reporting Style × Context × Word was not
significant, Fs < 2.9, ps > .09. It suggests that loudness (i.e., vocal volumes) may not be
encoded as part of the simulated „inner voice‟ experiences during silent reading of direct as
opposed to indirect speech, and that vocal loudness may be simulated in other types of
representations such as prosody. No other effect was significant except that the Reporting
Style × Context interaction was significant by subject, F1(1, 47) = 14.696, p < .001, and
marginally significant by item, F2(1, 23) = 4.030, p = .057. Exploring this interaction, we
found that the Context effects were not significant in the direct speech conditions whereas
during silent reading of indirect speech, the participants responded significantly faster
when reading the loud-speaking context than when reading the quiet-speaking context (see
Table 8). The significant between-context difference in the indirect speech condition may
reflect a „baseline‟ difference in emotional intensity across Context: The loud-speaking
stories are usually more intense and emotionally engaging (e.g., shouting) than the quietspeaking stories (e.g., speaking politely); hence participants tended to respond faster while
reading the former because they were more roused.
However, other latent factors
significantly diminished this „emotional intensity‟ RT difference in the direct speech
Page | 107
conditions, which resulted in the observed Reporting Style × Context interaction. Such
result patterns cannot be simply interpreted in mental simulations of „supra-modal‟
emotional representations. This is because that direct speech is represented in a more vivid
fashion than indirect speech; one would predict that individuals should be more likely (or
no less likely) to mentally simulate emotion representations during silent reading of direct
speech as opposed to indirect speech. If participants were more roused when reading loudspeaking indirect speech quotations than quite-speaking indirect speech quotations (as
suggested by the significant Context effect in the indirect speech condition), they should
have been even more roused or equally more roused during silent reading of loud-speaking
direct speech quotations as compared to quiet-speaking direct speech quotations. However,
the latter was clearly not the case: The RTs in the loud-speaking direct speech condition
was significantly slower than the loud-speaking indirect speech condition (Table 8), and
the Context contrast in the direct speech condition was not significant. Thus, latent factor
which could exert inhibitory effects on RTs when reading loud-speaking direct speech
quotations is needed to interpret the non-significant Context contrast in the direct speech
condition. Suggested by our previous findings, one candidate could be mental simulations
of voices. It has been evident that individuals mentally simulate voice representations
during silent reading of direct speech but not indirect speech.
The simulated voice
representations are very likely to be acoustically distinct from the voices of the spoken
probe words, even when they are congruent in loudness. This means that when our
participants hear spoken probe words during silent reading of direct speech, they had to
pay a cost to integrate the mentally simulated voices and the spoken words‟ voices before
making a judgement response. Such a „voice integration‟ cost was consistent with the
significant RT delay in the loud-speaking, direct speech condition as compared to the loudspeaking, indirect speech condition. However, such a RT delay was not so significant in
the quiet-speaking conditions. It perhaps suggests the reported speaker‟s loudness is
represented in the salience of the simulated voice representations. If the „quite‟ voice
representations are less salient than the „loud‟ voice representations, so would be the
discrepancy between the simulated „quiet‟ voices and the spoken words‟ voices. Thus,
participants would pay less cost to integrate the two voice representations when reading
quiet-speaking direct speech and this could explain why their RTs were not as much
delayed as compared to when reading loud-speaking direct speech. Taken together, the
Reporting Style × Context interaction could be interpreted as the combination of the
Page | 108
following effects (also see Figure 13 for an illustration): (1) The Context effects (i.e., faster
RTs for the loud-speaking context) in the indirect speech condition reflected a betweencontext difference in emotional intensity; (2) such emotional intensity effects were
diminished in the direct speech condition due to a more salient „voice integration‟ cost in
the loud-speaking context than in the quiet-speaking context.
Loudly-speaking Context
Quietly-speaking Context
t(60)
By
p
Context subject
Contrast
t(23)
By
item
p
Reporting Style Contrast
By subject
By item
t(60)
p
t(23)
p
3.94 < .001
2.31
.03
-1.15
.255
-.6
.55
Direct
Speech
Indirect
Speech
732 (203)
724 (202)
.77
.44
.51
.61
697 (176)
732 (203)
-3.45
.001 Note: Significant contrasts are
-2.77 highlighted in bold text
.01
Table 8 RT means (in ms, alongside with standard deviations in parentheses) on the critical direct
speech and indirect speech quotations across Context. Also shown are results of paired-sample t-tests
investigating the simple contrasts of Reporting Style and Context.
Figure 13 A potential interpretation of the Reporting Style × Context interaction in Experiment 8. The
circles represent hypothetical RT means for the four conditions. The top panel illustrates the hypothetical
result pattern when only the Context (emotional intensity) main effect is considered. The middle panel
illustrates the hypothetical result pattern when only the voice simulation effect (the Reporting Style × Context
interaction) is considered. The bottom panel illustrates the hypothetical result pattern when both the
Context main effect and the voice simulation effects are considered – it resembles the results we found in
Experiment 8.
Page | 109
In summary, the current data are in line with voice simulations during silent reading of
direct speech but not indirect speech. They suggest that the reported speaker‟s loudness
may not be encoded in the volume of his/her voice. Instead, it may be represented in terms
of the salience of the simulated voice representations.
Future research is needed to
validate this „salience‟ hypothesis of loudness representations. For example, we could
measure the amplitude and the distribution of the brain activity within voice-selective areas
in relation to loud- versus quiet-speaking contexts during silent reading of direct speech as
opposed to indirect speech.
5.4
General discussion
The current chapter investigated the loudness aspect of the simulated „inner voice‟
representations during oral reading (Experiment 7) and silent reading (Experiment 8) of
direct versus indirect speech.
Experiment 7 showed that individuals spontaneously adjusted their own loudness in
accordance with the contextually implied loudness of the reported speaker during oral
reading of direct as opposed to indirect speech. In line with our hypothesis, the results
could mean that individuals are more likely to mentally simulate the reported speaker‟s
loudness during oral reading of direct speech as opposed to indirect speech.
However, mental simulations of loudness were not straightforwardly evident in silent
reading. In Experiment 8, participants silently read story-embedded direct and indirect
speech and responded to spoken words the volume of which either matched or mismatched
the reported speaker‟s loudness that was linguistically described in the contexts.
Inconsistent with our prediction, the participants‟ RTs were not effectively modulated
either by the volume (intensity level) of the spoken words or by their congruency with the
contexts. This could suggest that the paradigm was not sensitive enough to detect the
simulated loudness representations because the probe word was presented when reading
the final word of the speech sentence where voice simulations may have weakened over
time. Future research should use a probe word that occurs early during reading of the
speech utterance as opposed to at the final word of the sentence. Alternatively, it could
suggest that the reported speakers‟ loudness may not be encoded in terms of the absolute
vocal volume. Such results are consistent with Pitt and Crowder‟s (1992) research on
Page | 110
auditory imagery of tone loudness. In their investigations, they examined the effects of the
subjectively-imagined loudness (instructed with written words „loud‟ or „soft‟) of a sine
wave tone on the subsequent pitch judgement on a probe tone which is either loudnesscongruent or loudness-incongruent. They did not observe any loudness-congruency effect,
either in RTs or in accuracy rates, suggesting that acoustic representations of loudness
(volume) may not be encoded in loudness images. The authors hence proposed that in
imagery, loudness may be encoded in another form (e.g., the prosodic information and the
articulatory motoric information). In line with this speculation, we speculated that the
simulated loudness in silent reading of direct versus indirect speech is also encoded in
other forms. One such form, suggested by the significant Reporting Style × Context
interaction observed in our silent reading investigation, may be the salience of the
simulated voice representations.
Another possibility, suggested by the contextual
modulation of loudness we observed during oral reading of direct rather than indirect
speech (Experiment 7), is that loudness may be encoded in mental simulations of loudnessrelated articulations, i.e., in a form of motor simulations.
Nevertheless, further
investigations are needed to verify how loudness is mentally represented during
comprehension of direct versus indirect speech, using more sensitive methodologies such
as fMRI.
Page | 111
Chapter 6 Contextual modulation of reading rates for direct
versus indirect speech quotations
6.1
Introduction
The present chapter addresses another acoustic aspect (in addition to pitch and loudness) of
the purported „inner voice‟ experience during reading of quotations, namely the
contextually implied rate of the quoted speech. Specifically, we examined whether readers
adjust their reading rates in accordance with the contextually implied speech rate of the
quoted speaker during both oral reading (Experiment 9) and silent reading (Experiment 10)
of direct versus indirect speech.
Moreover, by focusing on potential behavioural
consequences of voice simulations during reading, the present studies represent an
important extension of the brain-imaging research reported in Chapters 2 and 3.
It has been demonstrated that reading rates may reflect auditory imagery of speaking rates
(Alexander & Nygaard, 2008). In their study, participants were exposed to pre-recorded
voices of either fast or slow speakers. In subsequent reading trials, participants were
prompted to imagine those speakers as authors of the given reading materials. It was found
that both oral and silent reading rates were faster when participants imagined previously
heard fast speakers as authors.
Although related, the present experiments are somewhat different. First, Alexander and
Nygaard (2008) did not investigate any influences of reporting style (direct versus indirect
speech), which are of paramount interest in our investigations. Second, their experimental
setup may have encouraged auditory imagery during reading, focusing on how the auditory
memory of a specific talker‟s speaking rate modulates reading rates. In contrast, our
experiments will examine how linguistic context (implying either a fast-speaking or a
slow-speaking quoted protagonist) modulates participants‟ reading rates during oral and
silent reading of quotations. Just as in Yao et al. (in press), participants in our experiments
will neither be exposed to auditory samples of specific speakers before reading, nor be
instructed or encouraged to „act out‟ or imagine voices. Moreover, the reading materials
will only contain fictitious, unfamiliar speakers. Thus, our experiments are designed to tap
into more automatic voice simulation processes in response to written quotations. Inspired
by Alexander and Nygaard (2008), we assume that such automatic voice simulations
Page | 112
should be measurable in modulations of oral and silent reading rates, and informed by
Chapter 2, we predict that these reading rate adjustments should be stronger for direct
speech quotations than for meaning-equivalent indirect speech quotations.
In the following experiments, participants either orally (Experiment 9) or silently
(Experiment 10) read short fictitious stories that contained either direct or indirect speech
quotations. The quotations were embedded in linguistic contexts describing either a fast- or
a slow-speaking reported speaker.
Participants‟ oral and silent reading rates were
measured via audio recordings and eye-tracking, respectively.
6.2
Experiment 9
6.2.1 Participants
Twenty native English speakers with no reported reading impairments participated in
exchange for £2. A typical session lasted about 20 minutes.
6.2.2 Stimuli and design
A 2 [Reporting Style: direct versus indirect speech] × 2 [Speaking rate: fast- versus slowspeaking reported speaker] design was used (see an example below).
Twenty-four
quadruples of short fictitious stories were prepared as reading materials (see Appendix 4
for a complete stimulus set). The stories contained either a direct speech or an indirect
speech quotation from a fictitious main protagonist in the story (Reporting Style). Across
items, different protagonist names were used.
A second variable (Context) was
independently manipulated such that the quotations were preceded by a context that
implied either a fast-speaking or a slow-speaking quoted protagonist. Crucially, critical
quotation passages within each item (bold texts) were identical between the fast- and slowspeaking Context conditions, and nearly the same between the direct and indirect speech
Reporting Style conditions. The quotations were always followed by at least one additional
sentence.
Page | 113
Fast-speaking context
Slow-speaking context
Direct
speech
It was a typical British day, rainy and gloomy.
Sixteen year-old pianist Bobby was going to
play in the quarter-finals of a local talent
competition. He was extremely nervous before
his performance. His mother encouraged him
but he was all shaking and said: ‘No! I can’t do
it! This is the end of the journey because it is
unlikely that I will make it this time.’ His
mother tried to calm him down, saying that it’s
not the winning that counts, but the taking part.
It was a typical British day, rainy and gloomy.
At Glasgow Royal Infirmary, an old man was
dying, and too weak to sit up. His family
members were sitting around the bed, feeling
sad. He wanted to say something, so his
daughter placed a cushion under his head.
Slowly, he looked around and said: ‘I’m
grateful you’re all here. This is the end of the
journey because it is unlikely that I will
make it this time.’ Then he closed his eyes
and everyone burst into tears.
Indirect
speech
It was a typical British day, rainy and gloomy.
Sixteen year-old pianist Bobby was going to
play in the quarter-finals of a local talent
competition. He was extremely nervous before
his performance. His mother encouraged him
but he was all shaking and said that he couldn’t
do it and that it was the end of the journey
because it was unlikely that he would make
it this time. His mother tried to calm him down,
saying that it’s not the winning that counts, but
the taking part.
It was a typical British day, rainy and gloomy.
At Glasgow Royal Infirmary, an old man was
dying, and too weak to sit up. His family
members were sitting around the bed, feeling
sad. He wanted to say something, so his
daughter placed a cushion under his head.
Slowly, he looked around and said that he
was grateful for their coming and that it was
the end of the journey because it was
unlikely that he would make it this time.
Then he closed his eyes and everyone burst
into tears.
Four stimulus-lists with counterbalanced item-condition combinations (6 stories per
condition per list) were constructed and word-processed into four versions of written
scripts. Each item appeared in one condition per list, but in different conditions across lists.
The order of the stories per script was randomised. Each script was randomly assigned to
five participants.
6.2.3 Procedure
The experiment was conducted in a sound attenuated room. Participants were instructed to
read out the stories from the script in one go and as naturally and fluently as possible. The
instructions did not explicitly encourage participants to vocally „act out‟ the stories during
reading.
After setting up the microphone and digital voice recorder, the experimenter sealed the
room and the participant started reading the stories aloud while audio recordings were
taken. These were saved as wave-files for later analyses.
Page | 114
6.2.4 Results and discussion
All recordings were listened to individually by the experimenter. Around 5% of the trials
were excluded from analysis because of word repetition or substitution errors.
The
remaining valid recordings were digitally visualized (in millisecond resolution) using
GoldWave 4.26. For each trial, the temporal onsets and offsets of the critical quotation
passages were identified. Next, oral reading rates (in numbers of syllables per second) for
the critical quotations were calculated and submitted to 2 [Reporting Style: direct speech
versus indirect speech] × 2 [Context: fast versus slow] ANOVAs with subject (F1) and
item (F2) as random factors.
Apart from a Context main effect by subjects (F1 (1, 19) = 4.74, p < .05, F2 (1, 23) = 1.68,
p > .2) there was a significant Context  Reporting Style interaction (F1 (1, 19) = 8.90, p
< .01, F2 (1, 23) = 6.40, p < .02). Exploring the latter, it was found that direct speech
quotations were read out significantly faster when the context implied a fast-speaking
rather than a slow-speaking quoted protagonist.
By contrast, no such contextual
modulation of oral reading rate was found for indirect speech quotations (Table 9). This
suggests that participants engaged in spontaneous vocal re-enactments of the reported
speech act when reading aloud direct rather than indirect speech quotations, by adjusting
their oral reading rates to the speech rates that were contextually implied. Importantly, the
results cannot be explained by differences in length or morphological features (tense,
person, etc.) between the two Reporting Style conditions because the interaction involved
differential effects of Context within each Reporting Style condition.
Direct
Speech
Fast-speaking Context
Slow-speaking Context
t(19)
By
Subjects
p
Context
Contrast
t(23)
By
Items
p
6.17 (1.0)
5.88 (0.9)
3.65
.002
2.77
.01
Quoting Style Contrast
By Subjects
By Items
t(19)
P
t(23)
p
5.85 (0.9) 3.02 0.007
2.41
.03
5.93 (1.0) -0.37
.71 -0.46
.65
-0.37
.71 Note: Significant contrasts are
-1.13 highlighted in bold
.27
Indirect
Speech
Table 9 Oral reading rates (in syllables per second, with standard deviations in parentheses) for the
critical quotations in Experiment 9. Also shown are results from 2-tailed paired sample t-tests examining
the simple effects of Context and Reporting Style.
Page | 115
6.3
Experiment 10
Experiment 10 was an eye-tracking study using the same experimental items as
Experiment 9. The question was whether the results from Experiment 9 would replicate if
participants were reading the written materials silently for comprehension.
6.3.1 Participants
Fifty-two native English speakers with no reported reading impairments participated in
exchange for £3. A typical session lasted 30 minutes. Four participants were excluded
from analysis due to poor eye-tracking (3 participants) or less than 75% answering
accuracy (1 participant), leaving 48 participants for analysis.
6.3.2 Stimuli and design
The stimuli and design were identical to Experiment 9. The critical stories (24 items) were
mixed with 24 filler stories, which contained no direct or indirect speech quotations to
conceal the purpose of the study. Experimental and filler items spanned 4-7 lines on
screen. Although it was impossible to keep the formatting absolutely identical, care was
taken to ensure that the critical quotations spanned roughly the same screen areas across all
four conditions within each item.
Content-related questions were also prepared to
encourage reading for comprehension.
6.3.3 Procedure
The experiment was conducted using a SR-Research EyeLink 1000 desk-mounted eyetracker running at 1000 Hz sampling rate.
software
developed
at
the
University
Stimulus presentation was controlled by
of
Massachusetts
(EyeTrack
0.7.1;
www.psych.umass.edu/eyelab/software/). Participants were seated about 70 cm from a 21
inch CRT display running at 100 Hz refresh rate in 1280 × 960 pixel resolution. 3
characters equalled about one degree of visual angle. Materials were presented in an 18 pt
Bitstream Vera Sans Mono font printed in black on a light grey background. Line spacing
was set to 60 pt such that fixation locations could unambiguously be mapped onto a
corresponding line of text. Viewing was binocular, but only the participant's dominant eye
was tracked (as determined by a simple parallax test). A chin rest was used to keep
Page | 116
viewing distance constant and to prevent strong head movements during reading. Button
responses were collected using a hand-held Microsoft USB game pad.
Each participant was assigned to one of four lists containing different versions of
experimental items, pseudo-randomly interspersed with the filler stories.
There were
always two fillers at the beginning. At the start of the experiment, the standard EyeLink
calibration and validation procedure was performed in which participants had to look at
nine fixation targets in random succession.
Calibration and validation were repeated
during the experiment if the experimenter noticed a decline in measurement accuracy.
Each trial started with the presentation of a central fixation dot for drift-correction,
followed by a small rectangle in the same location as the first character of the upcoming
text display. A fixation for at least 200 ms on this rectangle triggered the presentation of
the text, so that reading always started in the first character position. Participants were
instructed to read the stories carefully and to press a button when they had finished reading.
A simple yes/no comprehension question was then presented, which participants had to
answer using either the left („yes') or the right („no‟) response button. Answering the
question trigged the presentation of the next trial.
6.3.4 Results and discussion
Fixation coordinates were mapped onto character positions using EyeDoctor 0.6.5
(www.psych.umass.edu/eyelab/software/).
Fixations below 80 ms were pooled with
temporally adjacent fixations if the latter were within half a degree of visual angle.
Fixations on the critical quotation regions were summarized in terms of (i) go-past time
(also known as regression path duration), i.e., the time from fixating the critical region for
the first time until a subsequent region is fixated, and (ii) no-regressions go-past time (the
same as before, but excluding trials where readers regressed back to an earlier region after
initial reading of the critical region).
Page | 117
Condition
Measures
Go-past reading
time (ms)
Number of go-past
fixations
Go-past fixation
duration (ms)
Table 10
Fast-speaking
Direct
Fast-speaking
Indirect
Slow-speaking
Direct
Slow-speaking
Indirect
1919 (805)
2099 (904)
2013 (853)
2059 (1001)
9.9 (3.9)
10.6 (4.2)
10.2 (4.0)
10.4 (4.4)
194 (34)
198 (31)
198 (32)
197 (31)
Standard eye movement measures of processing across 4 conditions (with standard
deviations in brackets).
The standard eye movement measures of processing (i.e., without removing „outlier trials‟)
across all 4 conditions were reported in Table 10. Outlier trials were removed in two steps.
First, we excluded trials where the initial reading of the critical region was not fluent,
operationally defined by a Fixation Density Disparity greater than two (see Appendix 5).
Next, trials with a go-past time of more than 3 SD above an item‟s condition mean were
removed. In total, 74 trials (6.4%, equally distributed across conditions) were excluded as
a result of these two outlier definitions.
Data from the remaining 1078 trials were
submitted to 2 × 2 ANOVAs by participants and items.
There was a significant Context  Reporting Style interaction in go-past time (F1(1, 47) =
22.79, p < .001, F2(1, 23) = 6.41, p < .02): Direct speech quotations were read significantly
faster when the context implied a fast-speaking rather than a slow-speaking protagonist,
while no such Context contrast was found for indirect speech quotations (Table 11). In noregressions go-past time (excluding 16% of trials where readers regressed back to an
earlier region) the same Context  Reporting Style interaction emerged (F1(1, 47) = 14.90,
p < .001, F2(1, 23) = 4.91, p < .04, Table 12). The latter confirms that the go-past time
effects in Table 11 were not driven by re-inspections of earlier regions.
The by-subject t-tests in Table 11 and Table 12 also suggest a „reversed‟ (fast > slow)
simple effect of Context in the indirect speech condition. Closer inspection revealed that
this contrast was due to only three stimuli (items 1, 23, and 24 in Appendix 4). After
removing those stimuli from analysis, the Context  Quoting Style interaction remained
significant (go-past time: F1(1, 47) = 7.50, p < .01, F2(1, 20) = 4.65, p < .05; noregressions go-past time: F1(1, 47) = 6.00, p < .01, F2(1, 20) = 3.90, p = .062), but
crucially, the Context effect in the indirect speech condition disappeared (Table 13 and
Table 14). It is therefore hard to interpret the latter as anything other than item-specific
Page | 118
idiosyncrasy. By contrast, the Context-effect in the direct speech condition (fast < slow)
robustly showed up in all analyses.
In conclusion, not only oral readers (Experiment 9), but also silent readers adjust their
reading rates to the contextually implied speech rate when reading direct speech as
opposed to indirect speech quotations.
Again, the observed interactions cannot be
explained by differences in length and/or morphology between the two Reporting Style
conditions. Additional analyses of numbers and average durations of go-past fixations
(Appendix 6) revealed that the reported go-past time effects were mainly carried by fewer
fixations being made in the fast-speaking conditions.
Average durations of go-past
fixations did not seem to be affected much by the experimental manipulations.
Direct
Speech
1857 (743)
Fast-speaking Context
2011 (830)
Slow-speaking Context
t(47)
-3.50
By
Subjects
p
.001
Context
Contrast
t(23)
-2.32
By
Items
p
.03
Quoting Style Contrast
By Subjects
By Items
t(47)
p
t(23)
p
2056 (885) -3.17
.003 -2.56
.02
1942 (896) 0.50
.62
0.29
.78
1.84
.07 Note: Significant contrasts are
1.43 highlighted in bold
.17
Indirect
Speech
Table 11 Go-past times (in ms, with standard deviations in brackets) for the critical quotations in
Experiment 10 (silent reading). Also shown are results from 2-tailed paired sample t-tests examining the
simple effects of Context and Reporting Style.
Direct
Speech
1784 (693)
Fast-speaking Context
1932 (799)
Slow-speaking Context
t(47)
-3.06
By
Subjects
p
.004
Context
Contrast
t(23)
-2.42
By
Items
p
.02
Quoting Style Contrast
By Subjects
By Items
t(47)
p
t(23)
p
1957 (853) -3.42
.001 -2.43
.02
1845 (820) 1.01
.32
0.04
.97
2.55
.01 Note: Significant contrasts are
1.16 highlighted in bold
.26
Indirect
Speech
Table 12 No-regressions go-past times (in ms, with standard deviations in brackets) for the critical
quotations in Experiment 10 (silent reading). Also shown are results from 2-tailed paired sample t-tests
examining the simple effects of Context and Reporting Style.
Page | 119
Direct
Speech
1919 (752)
Fast-speaking Context
2081 (828)
Slow-speaking Context
t(47)
-3.81
By
Subjects
p
.001
Context
Contrast
t(20)
-2.78
By
Items
p
.01
Quoting Style Contrast
By Subjects
By Items
t(47)
p
t(20)
p
2043 (897) -1.90
.06 -2.13
.05
2024 (915) 0.35
.73
0.63
.54
0.25
.80 Note: Significant contrasts are
0.37 highlighted in bold
.72
Indirect
Speech
Table 13 Go-past times (in ms, with standard deviations in brackets) for the critical quotations in
Experiment 10 (silent reading), after removing items 1, 23, and 24 from analysis. Also shown are results
from 2-tailed paired sample t-tests examining the simple effects of Context and Reporting Style.
Direct
Speech
1843 (701)
Fast-speaking Context
1996 (800)
Slow-speaking Context
t(47)
-3.27
By
Subjects
p
.002
Context
Contrast
t(20)
-2.80
By
Items
p
.01
Quoting Style Contrast
By Subjects
By Items
t(47)
p
t(20)
p
1941 (860) -2.14
.06
.04 -2.00
1904 (838) 1.45
.15
0.64
.53
0.85
.40 Note: Significant contrasts are
0.33 highlighted in bold
.75
Indirect
Speech
Table 14 No-regressions go-past times (in ms, with standard deviations in brackets) for the critical
quotations in Experiment 10 (silent reading), after removing items 1, 23, and 24 from analysis. Also
shown are results from 2-tailed paired sample t-tests examining the simple effects of Context and Reporting
Style.
Finally, we also performed by-item correlation analyses to evaluate the consistency of the
results between the two experiments. Indeed, the raw oral reading times (in ms) from
Experiment 9 correlated well with the silent reading data from Experiment 10 (r(94) = .86,
p < .001)14. Given that these correlations are partly due to length effects (longer quotations
take longer to read both orally and silently), we also performed multiple regression
analyses with silent reading times (Experiment 10) as criterion variables, and both length
(in numbers of words) and raw oral reading times (Experiment 9) as simultaneous
predictors. These analyses obtained partial correlations of r(93) = .56 (p < .001), which
means that the two experiments obtained very consistent by-item results even when length
effects were partialled out.
14
We considered both go-past and no-regressions go past times in these correlation analyses. Rounded to the
second decimal, the r-values were identical.
Page | 120
6.4
Summary and general discussion
In two experiments, we found that both oral and silent readers adjusted their reading rates
in accordance with the contextually implied speech rate of a quoted protagonist, but only
when quotations employed a direct rather than an indirect speech quotation style. In line
with our previous research using fMRI (Chapter 2), the current findings (particularly from
Experiment 10) support a hypothesised „inner voice‟ experience during silent reading of
direct speech quotations; more specifically, we showed that speech rate is an important
aspect of this inner voice experience and that mental simulations of voice are not only
reflected in brain-activation patterns and vocalisations but also in behavioural eyemovement patterns.
The present results also extended previous findings by Alexander and Nygaard (2008) who
found that pre-exposure to the speech rates of specific speakers influences how fast oral
and silent readers would process written stimuli that were supposedly authored by those
speakers. Specifically, the present experiments did not focus on this kind of speaker
adaptation. Instead, we were interested in how information provided by the linguistic
context (implying either a fast-speaking or a slow-speaking quoted protagonist) influences
participants‟ reading rates during oral and silent reading, and importantly, how different
reporting styles would affect such contextual modulations of reading rate. Another aspect
in which our studies differed from Alexander and Nygaard‟s (2008) – apart from using
eye-tracking instead of self-paced reading for silent reading – is that our experimental
instructions did not explicitly encourage participants to „act out‟ or imagine speech during
reading (see 6.1 Introduction). In Experiment 9 (oral reading) participants were just asked
to read the stories aloud and in Experiment 10 (silent reading) experimental tasks
emphasised reading for comprehension. The observed contextual modulations of reading
rate in response to direct speech quotations therefore suggest that, in this type of quotations,
readers routinely and automatically activate mental representations that link to how the
quoted speaker would sound like (in this case, how fast the quoted protagonist would
speak), which in turn have an effect on the rate of information uptake (silent reading) and
delivery (oral reading). In this respect, the present results are an important extension of our
previous findings (e.g., Chapter 2), who equally found that silent readers of direct speech
quotations routinely and automatically activate voice-related perceptual representations.
Page | 121
Our findings may be interpreted in terms of perceptual simulation, as proposed by
embodied cognition theories (Barsalou, 1999, 2008). Such theories argue that mental
representations of language are grounded in perceptual experiences and actions, and that
perceptual simulation (i.e., the mental re-enactment of perceptual, motor, and introspective
states acquired during experience with the world, body and mind) is an automatic and
integral part of language comprehension. Under such a premise, one could argue that
accumulated experiences with how direct versus indirect speech quotations are typically
used form the basis for differential degrees of perceptual simulation during language
comprehension. When speakers employ direct speech, they often mimic or dramatize
aspects of the reported speaker‟s voice in order to depict the reported speech act; indirect
speech, by contrast, is typically not used in such a demonstrative fashion (Clark & Gerrig,
1990). Comprehension of direct speech is therefore more likely to be grounded in the
perceptual experience of a vocal demonstration or dramatization of a reported speaker‟s
utterance, and thus more likely to invoke perceptual simulations of the reported speaker‟s
voice. Put differently, a direct-speech reporting style is more likely to be taken as a cue to
spontaneously engage in vivid perceptual simulations of the reported speech act than an
indirect-speech reporting style.
Of course, there are still a number of open questions which our research cannot answer at
present. Some are related to the nature of voice simulation itself. For instance, is the level
at which voice simulation operates primarily prosodic (especially given that loudness may
not be encoded in such processes; cf. Chapter 5), and what kind of processing takes place
in conditions where voice simulation is reduced or absent? Other questions relate to eyemovement control in reading. The results from Experiment 10 suggest that high-level
factors such as context and quoting style have an impact on basic reading rates. Do these
pragmatically induced reading-rate modulations operate at the level of individual word
processing or at the level of more „global‟ reading parameters? As these questions indicate,
further research into this area might substantially deepen our understanding of written
language processing as a whole.
In conclusion, our investigations address the pragmatics of different quotation styles and
their associated mental representations during text comprehension.
Our experiments
showed that oral and silent readers spontaneously adjust their reading rates to contextually
implied speech rates when reading direct speech as opposed to meaning-equivalent indirect
Page | 122
speech quotations.
These reading-rate adjustments may be interpreted as perceptual
simulations of reported speakers‟ voices.
Specifically, it suggests that temporal
representations such as speaking rate are an important aspect of such simulated „inner
voice‟ experience in reading direct speech as opposed to indirect speech.
Page | 123
PART 3: The Visual Aspect of the Direct-Speech Representations:
Mental Simulations of Facial Expressions?
Page | 124
Chapter 7 Are facial expressions also encoded in silent reading
of direct speech? (Experiment 11)
7.1
Introduction
Using fMRI, we have shown that individuals are more likely to mentally simulate the
reported speakers‟ voices during language comprehension of direct as opposed to indirect
speech (Chapter 2); the simulated voice representations entail supra-segmental acoustic
information (Chapter 3). At the behavioural level, we found that the simulated voice
representations could be characterised in modulations of acoustic parameters such as
speaking rate. The findings are in line with Clark and Gerrig‟s (1990) demonstration
theory of direct speech and the embodied theories in language processing (e.g., Barsalou,
1999, 2008; Zwaan, 2004).
However, in verbal communication, direct-speech demonstrations usually entail not only
vivid depictions of the reported speakers‟ voices but also depictions of other non-vocal
paralinguistic information such as the reported speaker‟s facial expressions. Hence, a
thorough investigation of the mental representations during silent reading of direct versus
indirect speech should, in addition to voices, demonstrate simulations of other non-voice
aspects such as facial expressions. The literature on recognition of facial expressions
suggests that understanding facial expressions involves embodied simulations of facial
expressions by the perceivers themselves (see Niedenthal, Mermillod, Maringer, & Hess,
2010 for a review and the proposal of the Simulation of Smiles (SIMS) model). Such a
„simulation‟ mechanism of understanding facial expressions has recently been
demonstrated in language comprehension (Havas, Glenberg, Gutowski, Lucarelli, &
Davidson, 2010). The authors applied subcutaneous injections of botulinum toxin-A (BTX)
to temporarily paralyse participants‟ facial muscles and asked them to read sentences (e.g.,
The pushy telemarketer won‟t let you return to your dinner) implying certain emotions
(e.g., anger). They found that the BTX treatment selectively slowed down the reading of
sentences that described situations which require the paralysed muscle for expression the
emotions evoked by the sentences. Their results indicate that processing of emotional
language requires embodied simulations of facial expressions and causally activates facial
muscles that are used in expressing the relevant emotions. Regarding comprehension of
direct versus indirect speech, re-analysis of our previous fMRI results (Chapter 2) revealed
Page | 125
that silent reading of direct speech elicited higher brain activations not only within the
voice-selective areas of the auditory cortex but also within the occipital cortex, superior
parietal lobules and precuneus (Figure 14A). Activation of these non-voice areas is in line
with an enriched multi-sensory mental simulation process for direct speech that also
encompasses, for instance, visual aspects of the described situation. When we applied a
face localiser15 on the whole-brain contrast, we observed that the brain areas that are more
active when seeing face images as opposed to non-face images (e.g., the bilateral fusiform
face areas (FFA); cf. Kanwisher, McDermott, & Chun, 1997, and the right posterior
superior temporal sulcus (pSTS)) were also more active during silent reading of direct
speech as opposed to indirect speech (Figure 14B). The findings indicate the possibility
that individuals may indeed mentally simulate face representations during silent reading of
direct speech but not indirect speech.
Figure 14 The brain activations in the contrast of direct speech as opposed to indirect speech in
Experiment 1 (Chapter 2). A, the whole-brain contrast, B, the between-condition differences masked by a
face localiser (the green areas indicate the brain areas that are more sensitive to seeing face images than to
seeing non-face images). Only the group ROI is used because the subjects were from an independent sample.
15
The face localiser was obtained from a different sample of 18 participants. During the localiser task,
participants were instructed to look at images that are presented at the centre of the screen. The 5 20-sec
blocks of faces, 5 20-sec blocks of face-like houses and 5 20-sec blocks of scrambled visual patterns were
presented in an efficiency optimised, pseudo random order. The face-sensitive brain areas were obtained by
contrasting the face condition to both the house condition and the visual pattern condition.
Page | 126
Thus, in this chapter, we ventured beyond the „inner voice‟ experiences by investigating
whether individuals are more likely to mentally simulate reported speakers‟ facial
expressions during silent reading of direct as opposed to indirect speech. Since facial
expressions are important vehicles of emotions in human interactions (Matsumoto, Keltner,
Shiota, O'Sullivan, & Frank, 2008; Russell, Bachorowski, & Fernandez-Dols, 2003), we
manipulated the types of facial expressions by manipulating the reported speakers‟
emotions. We examined three emotions, joy, surprise and anger, which are of similar
arousal levels and are of either positive or negative valence; they would presumably cover
the range of facial expressions in a relatively comprehensive way.
We adopted the
„reading & recognition‟ paradigms in the literature on visual simulations (e.g., Connell,
2007; Stanfield & Zwaan, 2001; Zwaan et al., 2002). Specifically, we examined the
effects of the linguistically implied facial expressions of the reported speakers on the
recognition latencies to visual stimuli of facial expressions that were made relevant by the
previous text passages. We prepared short stories containing direct or indirect speech in
which the reported speakers were either happy, or surprised, or angry. And we prepared
graphic emoticons which illustrate facial expressions of the investigated emotions as the
visual probe stimuli to match the stories. We chose graphic emoticons over photorealistic
face images as the visual probe stimuli because the former are highly abstract in the
confounding identity-specific information such as the person‟s age, gender, skin type, etc.
that we were not interested in, and yet they still illustrate sufficiently detailed facial
expressions and have been found to activate face-sensitive brain areas such as the fusiform
face areas (FFA) in a similar manner as the latter (Yuasa, Saito, & Mukawa, 2011).
Participants silently read a series of short stories containing direct or indirect speech, each
followed by a graphic emoticon, and answered „yes‟ or „no‟ to indicate whether the
illustrated expression was congruent with the reported speaker‟s facial expression that was
implied in the preceding text passage. If individuals mentally simulate reported speakers‟
facial expressions during silent reading of direct speech rather than indirect speech, their
recognition on expression-congruent graphic emoticons should be facilitated whereas their
recognition on expression-incongruent graphic emoticons should be inhibited, but only
after reading direct speech not indirect speech.
In addition, we manipulated linguistic focus on reported speakers‟ facial expressions to
verify the nature of the underlying representations of the purported differential face
Page | 127
simulations during silent reading of direct versus indirect speech. Similar to the Linguistic
Focus manipulation on voices in Experiment 3, we employed face-focused versus faceunfocused linguistic descriptions before the direct and indirect speech quotations (e.g.,
Vicky smiled jubilantly: „…‟ versus Vicky felt rather relieved: „…‟). Advised by the results
in Experiment 3, we predicted any observed differential face congruency effects during
silent reading of direct versus indirect speech should be selectively masked by exerting
linguistic focus on faces (see the masking hypothesis in 4.1 Introduction and see 4.2.5 for
the evidence of the masking hypothesis on voice simulations). In other words, facefocused descriptions should activate face representations (e.g., smiling faces from Vicky
smiled jubilantly) which would be integrated during silent reading of both direct and
indirect speech. Thus, we should observe a face congruency main effect regardless of the
reporting styles. By contrast, the purported differential face simulations in silent reading
of direct versus indirect speech should be retained when the quotations are preceded with
face-unfocused descriptions which do not activate face representations. And we should
observe an interaction between the face congruency factor and the reporting style factor.
In sum, if we observe that the differential face congruency effects during silent reading of
direct versus indirect speech can indeed be selectively masked by exerting the linguistic
focus on faces, we can conclude that the representations underlying such congruency
effects should contain face representations.
7.2
Participants
Sixty-four native speakers of English participated in this experiment (21 male, 43 female,
age 17 – 37, M = 21.6, SD = 3.5). They had no reported reading or relevant cognitive
impairments. They were paid £3 for their participation. A typical session lasts about 25
min.
7.3
Stimuli and design
A 2 [Reporting Style: direct versus indirect speech] × 2 [Face Match: the reported
speaker‟s facial expression matches versus mismatches the facial expression of the
emoticon] × 2 [Linguistic Focus: face-focused versus face-unfocused pre-quotation
descriptions] design was employed (see below for an example).
Page | 128
Story
In her dinner party, Vicky was a bit nervous because
she had prepared some exotic recipes that she had
never attempted before. Unexpectedly, all dishes
went extremely popular and some guests even
craved more after they finished.
Face
Focused
Face
Unfocused
Direct
speech
Vicky smiled jubilantly: ‘Wow, I didn’t expect that. I
am so glad that you all enjoyed it!’
Indirect
speech
Vicky smiled jubilantly, saying that she did not
expect that and she was so glad that they all
enjoyed it.
Direct
speech
Vicky felt rather relieved: ‘Wow, I didn’t expect that.
I am so glad that you all enjoyed it!’
Indirect
speech
Vicky felt rather relieved, saying that she did not
expect that and she was so glad that they all
enjoyed it.
Probe Picture
Face
Match
Face
Mismatch
Thirty-six short stories with different protagonists were prepared as reading materials. The
stories described protagonists who were either happy (12 stories), or angry (12 stories), or
surprised (12 stories).
Half of the „surprised‟ stories described positive events (e.g.,
finding an antique) whereas the other half described negative events (e.g., receiving news
of death). Together with the „happy‟ and „angry‟ stories, half of the 36 critical stories
described reported speakers of positive emotions while the other half reported speakers of
negative emotions. Each story started with two declarative sentences to set up a scenario,
followed by either a direct speech or an indirect speech from a main protagonist in the
story (Reporting Style factor).
The quotation was preceded by either a face-focused
description or a face-unfocused description (Linguistic Focus factor).
Three graphic
emoticons illustrating happy, surprised and angry emotions were created to accompany the
stories (see Figure 15); they either matched or mismatched the reported speaker‟s facial
expressions that were linguistically implied in the story (Face Match factor). For each
emotion, there was equal number of two mismatch emoticons: For instance, half of the
„happy‟ emotion items had the „surprised‟ emoticon as the mismatch stimulus whereas the
other half of the items had the „angry‟ emoticon as the mismatch stimulus. The probe
pictures were sized 237 × 237 pixels (ca. 9 inches2) and the graphic emoticons were of the
same colour, size and stylistic style.
Yes/no comprehension questions regarding the
contents of the stories (e.g., Was Vicky's dinner party successful?) were also prepared to
Page | 129
assess participants‟ overall comprehension performance and to ensure that they read the
stories attentively.
In addition to the 36 critical items, 42 filler stories were prepared to conceal the intended
experimental manipulations. They were accompanied with pictures of non-face objects
(e.g., a book) which were either congruent or incongruent with the corresponding stories.
Nine of the filler stories contained direct speech quotations and nine contained indirect
speech quotations. The remaining 24 filler items were declarative discourses with no
reported speech. In total, there were 54 stories which contained direct or indirect speech;
36 of them were accompanied with graphic emoticons (critical items) while 18 of them
accompanied with pictures of non-face objects. Yes/no comprehension questions were
also prepared for filler items.
Eight lists of counterbalanced item-condition combinations were yielded.
Each item
appeared in one condition per list and in different conditions across lists. Each participant
only saw one list. The presentation order was randomised for each participant.
Figure 15 Graphic emoticons illustrating ‘happy’, ‘angry’ and ‘surprised’ expressions. They were of the
same size, colour and style.
7.4
Procedure
Each participant was presented with one of the eight stimulus lists using E-prime 2.0
presentation software (Psychology Software Tools, Inc., USA).
In each trial, the
participant first saw a cue „NEW TRIAL‟ in the middle of the computer screen for 1000
ms, which was followed by the presentation of a short story. The text was presented in an
18 pt Calibri font on a light grey background. The whole text passage was centred and
wrapped within 65% width of the screen, together making the reading stimuli appear as
natural as possible. Participants were instructed to silently read the text passage in their
own time, and press spacebar to terminate the presentation as they finished reading. Next,
Page | 130
they saw a question mark at the centre of the screen for 1000 ms, followed by a graphic
emoticon. They were instructed to indicate as quickly and accurately as possible whether
or not the illustration (of either a facial expression or an object) was congruent with the
story they had read, by pressing mouse buttons (left for „yes‟, right for „no‟). At the end of
each trial, participants were prompted to answer a yes/no comprehension question
regarding the story they had just read.
7.5
Results and discussion
The participants‟ accuracy on the comprehension questions was all above 75% (M = 89%,
SD = 31.4%), indicating that they read the stories attentively. Surprisingly, the participants‟
accuracy on the emoticon recognition amounted to a mere 87% (SD = 34%), which was
even lower than the accuracy on the comprehension questions.
This figure was
unexpectedly low because previous investigations using similar „reading & recognition‟
paradigms (e.g., Stanfield & Zwaan, 2001) usually report a recognition accuracy of higher
than 95%. Thus, to ensure that our experimental manipulations were effective as intended,
we needed to identify the source and the cause of the recognition errors before we proceed
to test our hypothesis. To this end, we first conducted a logit binomial GEE (Hardin &
Hilbe, 2003) on participants‟ recognition accuracy (as the binary dependent variable), with
Reporting Style, Context Emotion (happy, surprised, and angry), Emoticon (happy,
surprised and angry) and Linguistic Focus as predictors. There was a significant Context
Emotion main effect, Generalised Score Chi-Square(2) = 25.48, p < .001, a significant
Emoticon main effect, Generalised Score Chi-Square(2) = 21.89, p < .001, a significant
Context Emotion × Emoticon interaction, Generalised Score Chi-Square (4) = 47.47, p
< .001, and a significant Context Emotion × Emoticon × Reporting Style × Linguistic Focus
interaction, Generalised Score Chi-Square(4) = 11.60, p = .021. The results indicate that
the participants‟ recognition errors were unevenly distributed across the Context Emotion
and the Emoticon manipulations. To illustrate these distributions, we summarised the
recognition error rates by these two factors in Table 15, alongside with their corresponding
RTs reported in Table 16. We found that the recognition errors were distinctly made in the
„surprised‟-related conditions, irrespective of whether the emotion was conveyed by the
linguistic context or by the emoticon. Consistently, the RTs in these „surprised‟ conditions
were also much longer than the „surprise-free‟ conditions. Such error distribution patterns
suggest that our participants had difficulty in processing the „surprised‟ expression which
Page | 131
they tended to confused with the „happy‟ and the „angry‟ expressions. Such confusions
indicate that the Face Match manipulations in the „surprised‟-related trials were not as
effective as intended. For this reason, we had to exclude all „surprised‟-related trials in
order to examine the effects of Face Match, which only left us 12 out of the 36 items for
the following analyses.
The mean error rates (%)
Emoticon
Happy
Surprised
Angry
Mean
Happy
1.04 (10.17)
34.90 (47.79)
0 (0)
9.24 (28.98)
Context Emotion
Surprised
Angry
3.65 (18.79)
42.19 (49.51)
10.94 (31.25) 21.88 (41.45)
5.21 (22.25)
22.92 (42.14)
21.74 (41.28)
8.98 (28.61)
Mean
11.98 (32.49)
19.66 (39.77)
8.33 (27.66)
Table 15 Mean error rates (in %, along with standard deviations in parenthese) across the Context
Emotion conditions and the Emoticon conditions. Unexpectedly high error rates (> 10%) were highlighted
in bold text.
The RTs (ms)
Emoticon
Happy
Surprised
Angry
Mean
Happy
671 (341)
1096 (656)
770 (320)
802 (470)
Context Emotion
Surprised
899 (701)
966 (652)
952 (527)
946 (636)
Angry
800 (574)
1144 (686)
863 (481)
913 (578)
Mean
755 (521)
1043 (665)
862 (463)
Table 16 Mean RTs (in ms, along with standard deviations in parenthese) across the Context Emotion
conditions and the Emoticon conditions. Also highlighted are the RTs in the conditions with high error
rates (see Table 15).
Incorrect responses to the spoken words were discarded from the valid 12 items (2.5%).
The outliers were excluded in two steps: First, the correct responses with a RT of longer
than 3000 ms or shorter than 300 ms were removed. Next, the remaining trials with a RT
which was 3SD away from an item‟s condition mean were excluded. In total, 4.8% (37 out
of 768 trials) of the trials were excluded from analyses.
The remaining trials were submitted to 2 [Reporting Style: direct versus indirect speech] ×
2 [Linguistic Focus: face-focused versus face-unfocused pre-quotation descriptions] × 2
[Face Match: the reported speaker‟s facial expression matches versus mismatches the
facial expression in the emoticon]16 ANOVAs with subject (F1) and item (F2) as random
factors, respectively.
We found that the Reporting Style × Face Match interaction
(ps > .26) and the Linguistic Focus × Reporting Style × Face Match interaction (ps > .9)
16
We merged the Context Emotion factor and the Emoticon factor into the Face Match factor to simplify the
interpretations of the effects.
Page | 132
were not significant. However, considering the inevitable reduction in statistical power
after discarding two third of our data, we nonetheless explored the patterns of these
interactions to assess whether they showed trends that were consistent with our hypothesis.
As for the Reporting Style × Face Match interaction, we found that after reading direct
speech, the participants responded 44 ms faster to the face-match emoticons as opposed to
the face-mismatch emoticons whereas in the indirect speech conditions, the RT difference
across the Face Match conditions was much less robust and in the opposite direction (the
match condition was 7 ms slower than the mismatch condition; see Table 17). This result
pattern was consistent with our hypothesis that individuals are more likely to mentally
simulate the reported speakers‟ facial expressions in silent reading of direct as opposed to
indirect speech. As for the Linguistic Focus × Reporting Style × Face Match interaction,
we split the data by Linguistic Focus and conducted two separate Reporting Style × Face
Match ANOVAs upon the face-focused data set and the face-unfocused data set,
respectively.
We found that when the speech quotations were preceded with face-
unfocused descriptions, the Reporting Style × Face Match interaction was significant by
item, F2(1, 11) = 6.89, p = .023, and was approaching significance by subject, F1(1, 52) =
2.41, p = .127. In the direct speech condition, RTs were significantly faster (by 97 ms;
significant by subject and marginal by item) in the face-match condition than in the facemismatch condition; whereas in the indirect speech condition, there was virtually no
difference (amounting to a mere 1 ms) between the face-match and -mismatch conditions
(see the upper panel in Table 18). As for the face-focused trials, such interaction was
„masked‟ (all ps > .6); the RTs in the face-match condition was numerically slower (by on
average 12 ms) than in the face-mismatch condition, independent of the reporting style
factor. Consistent with our predictions, the results indicate that the Reporting Style × Face
Match interaction can be selectively masked by exerting linguistic focus on faces. This
„face-selective‟ masking effect suggests that this interaction reflects mental simulations of
facial expressions (as opposed to non-face representations) during silent reading of direct
rather than indirect speech. When reconciled with the face-related brain activations in
silent reading of direct rather than indirect speech (Figure 14), the current findings
(although some results were not significant due to the exclusion of „surprised-related‟ trials)
indicates the possibility that individuals may indeed mentally simulate facial expressions
of the reported speakers during silent reading of direct rather than indirect speech.
Page | 133
Thus, future research would be needed to replicate the current result patterns, with three
aspects of experimentation improved. First, given that the „surprise‟ facial expressions are
not so clearly defined (both in terms of emoticons and in terms of simulations), the
experimental stimuli should perhaps only contain stories describing „happy‟ and „angry‟
emotions. Second, the current experimental setup required participants to press the left
mouse button for “yes” and the right mouse button for “no”. It is known that positive
emotions are typically (for right-handers) associated with rightward space whereas
negative emotions with leftward space (Daniel, 2009; Holmes & Lourenco, 2011). Hence
in our future research, we will counterbalance the button settings (i.e., half of the
participants press the left mouse button for “yes” while the other half press the right mouse
for “yes”) to prevent potential „button-pressing‟ bias towards emotionally-valenced stimuli
used in our investigations. Finally, we will also norm our reading materials in terms of
semantic equivalence to ensure that the Linguistic Focus manipulation is not confounded
with semantic discrepancies between the Focus and the non-Focus conditions.
Direct
Speech
Face Match
Face Mismatch
t(63)
By
Subjects
p
Context
Contrast
t(11)
By
Items
p
691 (262)
735 (232)
-1.66
.10
-.798
.44
Quoting Style Contrast
By Subjects
By Items
t(63)
p
t(11)
p
741 (306) -1.56
.12 -1.39
.19
734 (218) .165
.87
.117
.91
.21
.84 Note: Significant contrasts are
.21 highlighted in bold
.84
Indirect
Speech
Table 17 Mean RTs (in ms, along with standard deviations in parenthese) in the face match and face
mismatch conditions across Reporting Style. Also shown are the paired-sample t-test results examining the
simple effects of Reporting Style and Face Match.
Page | 134
Face-unfocused trials
(e.g., Vicky felt rather relieved)
Face Match
Face Mismatch
By
Subjects
Context
Contrast
By
Items
t(63)
p
t(11)
p
Face-focused trials
(e.g., Vicky smiled jubilantly)
Face Match
Face Mismatch
By
Subjects
Context
Contrast
By
Items
t(63)
p
t(11)
p
Direct
Speech
651 (244)
748 (237)
-2.51
.015
-2.02
.069
Direct
Speech
732 (273)
722 (228)
-.04
.97
.19
.85
Quoting Style Contrast
By Subjects
By Items
t(63)
p
t(11)
p
715 (298) -1.01
.32 -2.08
.06
716 (185) 2.24
1.12
.29
.03
.42
.68 Note: Significant contrasts are
.046 highlighted in bold
.96
Quoting Style Contrast
Indirect
By Subjects
By Items
Speech
t(63)
p
t(11)
p
766 (313) -.78
.44
-.65
.53
752 (246) -.13
.90
-.88
.40
.45
.66 Note: Significant contrasts are
.29 highlighted in bold
.78
Indirect
Speech
Table 18 Mean RTs (in ms, along with standard deviations in parenthese) in the face match and the
face mismatch conditions across Reporting Style in the face-unfocsed and the face-focused trials,
respectively. Also shown are the paired-sample t-test results examining the simple effects of Reporting Style
and Face Match in the two subsets of data.
Page | 135
Chapter 8 Conclusions and closing remarks
The experiments discussed in this thesis provide empirical support for the demonstration
theory of direct speech (Clark & Gerrig, 1990) and embodied theories of language
comprehension (e.g., Barsalou, 1999). In these experiments, we examined the mental
representations during language comprehension of direct versus indirect speech.
Specifically, we predominantly investigated whether individuals would be more likely to
mentally simulate the reported speakers‟ voices during comprehension of direct speech as
opposed to indirect speech as well as the nature of the simulated voice representations.
Moreover, we explored whether representations of facial expressions are also mentally
simulated during silent reading of direct versus indirect speech.
Experiments 1 and 2 (Chapters 2 and 3) obtained neuroimaging evidence of the
hypothesised mental simulations of voices during language comprehension of direct speech
rather than indirect speech. Experiment 1 demonstrated that silent reading of direct speech
elicited higher brain activity within the voice-selective areas of the right auditory cortex as
compared to silent reading of indirect speech.
In the absence of external auditory
stimulation, the observed brain activity within the voice-selective areas of the auditory
cortex was interpreted as „top-down‟ mental simulations of voices.
Experiment 2
demonstrated that virtually the same brain areas of the auditory cortex became more active
while listening to monotonously spoken direct as opposed to indirect speech.
Such
differential brain activations could not be accounted for by the acoustic characteristics or
the vividness of the speech utterances without context other than the perceived congruency
of the speech utterances in consideration of linguistic context. The findings suggest that
listeners routinely expect more vivid vocal depictions for direct than for indirect speech.
The findings are consistent with the notion that listeners spontaneously simulate suprasegmental acoustic representations when listening to the former in which such acoustic
information is minimised.
The observed brain activations in both experiments were
located within virtually the same brain areas, including the posterior, middle and the
anterior parts of the right superior temporal sulcus (STS). Taken together, we conclude
from the two studies that language comprehension of direct speech is grounded in mental
simulations of supra-segmental acoustic representations of voices whereas language
comprehension of indirect speech is not, and that the reporting style of direct speech is
intrinsically more vivid than that of the indirect speech. We also propose that the posterior,
Page | 136
middle and anterior parts of the right STS brain areas are involved in simulating such
supra-segmental acoustic representations.
Experiments 3 to 10 (Chapters 4 to 6) sought to determine the behavioural consequences of
the observed voice simulations during comprehension of direct versus indirect speech.
These experiments examined contextual modulation of the acoustic dimensions (including
pitch, loudness and rate) of the simulated voice representations during comprehension of
direct and indirect speech. We found that oral readers spontaneously adjusted their own
vocal acoustics (i.e., pitch, loudness or rate) in accordance with those of the reported
speakers that were linguistically implied in the preceding contexts only when reading
direct speech.
The findings suggest that the simulated voice representations in
comprehension of direct rather than indirect speech may be characterised in acoustic
parameters of pitch, loudness and rate, and that voice simulations are not only reflected in
brain-activation patterns but also in articulation. Moreover, we found that silent readers
also adjusted their reading rates in accordance with the contextually implied speaking rates
of the reported speakers when reading direct speech but not indirect speech. The results
verify that speaking rate is indeed an important acoustic aspect of the simulated voice
representations in comprehension of direct rather than indirect speech, and that voice
simulations can also be reflected in eye-movement patterns.
However, it remained
inconclusive whether and how pitch and loudness are encoded in voice simulations during
silent reading of direct rather than indirect speech. This could be because that the probe
word was presented when reading the final word of the speech sentence (where voice
simulations may have weakened over time), rendering the paradigm not sensitive enough
to detect the simulated pitch or loudness representations. Future research should use a
probe word that occurs early during reading of the speech utterance as opposed to at the
final word of the sentence. Overall, we may conclude that the voice representations that
are simulated in comprehension of direct rather than indirect speech can be characterised in
modulations of speaking rate while further investigations are needed to illuminate how
pitch and loudness are encoded in voice simulations.
To demonstrate a multi-sensory representation of direct speech, Experiment 11 (Chapter 7)
ventured beyond voice simulations by examining whether individuals also mentally
simulate, in addition to voices, the reported speakers‟ facial expressions during silent
reading of direct rather than indirect speech. We assessed the effects of the implied facial
Page | 137
expressions of the reported speakers on the recognition latencies to graphic emoticons in
which the illustrated facial expressions were made relevant by the previous text passage.
We investigated three facial expressions but found that the „surprised‟ expression was
frequently confused with the „happy‟ and „angry‟ expressions. After excluding „surprised‟related trials, we observed that when the linguistic focus was not on facial expressions,
readers responded significantly faster to expression-congruent graphic emoticons than to
expression-incongruent graphic emoticons after reading direct rather than indirect speech.
This interaction was selectively masked when linguistic focus was placed on facial
expressions, confirming that it reflected mental simulations of facial expressions (as
opposed to non-face representations) during silent reading of direct rather than indirect
speech. Future research is needed to replicate the current findings with more stimuli,
focusing only the „happy‟ and the „angry‟ expressions.
In conclusion, this doctoral research showed that individuals are more likely to mentally
simulate enriched supra-segmental acoustic representations of reported speakers‟ voices
during language comprehension of direct speech as opposed to indirect speech. One aspect
of such simulations may be characterised in modulations of speaking rate, but it remains
unclear whether they encode acoustic representations of pitch or loudness. Moreover, it
showed that representations of facial expressions may also be mentally simulated during
silent reading of direct rather than indirect speech, as part of a multi-sensory representation
of the former. The findings support the demonstration theory of direct speech (Clark &
Gerrig, 1990) from the perspective of language comprehension, suggesting that the direct
speech reporting style is intrinsically more vivid than the indirect speech reporting style.
The results also extend embodied theories of language comprehension (e.g., Barsalou,
1999) in several respects. First of all, by demonstrating mental simulations of voices in
language comprehension, this research extends the empirical testing of embodied theories
to the auditory perceptual modality and to the sentence/discourse level which so far has not
been thoroughly investigated in the relevant literature (previous studies have only focused
on sound-related words, see Kellenbach et al., 2001; Kiefer et al., 2008). In this way, it
contributes to the demonstration of the full spectrum of embodied phenomena in language
comprehension.
Second, while the contemporary empirical foundation of embodied
language comprehension focuses on the grounding of the semantic meanings of language,
the current research provides the first demonstration that language pragmatics, (e.g., the
Page | 138
vividness distinction between semantically-equivalent direct and indirect speech reporting
styles) underlie language comprehension: In verbal communication, direct speech usually
coincides with vivid paralinguistic demonstrations of the reported speech act whereas
indirect speech is reported in a less vivid fashion. This research shows that this vividness
distinction is also reflected in how language is processed, and that direct speech is mentally
represented in vivid paralinguistic demonstrations and is more likely to be associated with
mental simulations of voices than indirect speech. Third, this research combines eventrelated fMRI with eye-tracking to investigate neural correlates during on-line language
comprehension; it revealed that the posterior, middle and anterior parts of the right STS are
involved in mental simulations of supra-segmental acoustic representations. Such methods
could inspire new applications in psycholinguistic research on language comprehension,
helping us to understand the kinds of mental representations activated during reading.
At the end of the thesis, let us take a look into the future of embodied theories and their
applications in language research. Embodied theories have marked a fundamental change
in our current thinking about cognition by hypothesising that language, as well as other
high-level cognitive functions, are „embodied‟ in the sense that they are grounded in our
sensory and motor experiences (implemented via some form of simulations/re-enactments)
rather than in amodal symbols (e.g., Barsalou, 1999). However, contemporary empirical
testing has been limited to demonstrating embodiment-type effects (e.g., interaction
between linguistic and motoric tasks) in a rather superficial manner. To extend the scope
of this empirical demonstration of embodied language phenomena, the current doctoral
research combines both the traditional and the state-of-the-art neuroimaging methodologies
and probes into both the behavioural consequences and the neural basis of the purported
voice simulation processes. Nevertheless, this research cannot answer whether embodied
simulations are functionally necessary or causal to language comprehension, or whether
they play a relatively peripheral, or even epiphenomenal, role in language comprehension.
Nor does it address how abstract language (e.g., the concept democracy) is mentally
represented. These two issues (i.e., the necessity/causality of simulations in language
comprehension and the representational nature for abstract language use) are often used to
serve the argument against embodied theories, given that they have been the central issues
in the development of embodied theories and the limitation of the contemporary empirical
testing of these theories (Barsalou, 2010; Kiefer & Pulvermüller, in press). That said, the
Page | 139
current doctoral research provides a promising platform to address these issues. For
example, to assess the necessities of voice simulation in comprehension of direct vs.
indirect speech, we can use repetitive transcranial magnetic stimulation (rTMS) to
temporarily disrupt the functions of the „voice simulation‟-related brain areas. The effects
of such rTMS will implicate the causal role of voice simulations in language
comprehension of direct vs. indirect speech.
To explore the abstraction of voice
simulations, one may be interested in investigating how deaf people, who have no
experience of voice, mentally represent voices during comprehension of direct vs. indirect
speech.
Based on the findings obtained within this doctoral thesis, future research has the potential
to venture beyond demonstrations of embodied phenomena by seeking the evidence
necessary to develop explicit mechanistic models of embodiment, using a variety of
techniques
such
as
eye-tracking,
fMRI,
diffusion
tensor
imaging
(DTI),
magnetoencephalography (MEG) and rTMS. The combination of these techniques can
help us, for example, delineate the neural circuitry underlying simulations during written
discourse comprehension, verify the necessity of voice simulations and explore the
abstraction of voice simulations in the deaf population. The precise neural dynamics and
verified properties of simulations will lay the biological and empirical foundation for
sophisticated computational modelling of language comprehension and cognition, leading
to development of fully intelligent systems and effective robots.
Page | 140
Appendices
Appendix 1(Stimuli for Experiment 1 & 2)
[D] = direct speech; [I] = indirect speech; [Q] = comprehension question
1. Julie and Mark had been classmates and have not seen each other for years. Today, they met in the local
supermarket and Julie started a conversation about career paths.
[D] She said: „My life has been amazing! After merely three years, I‟m now a solicitor.‟
[I] She said that her life had been amazing, and that after merely three years, she now was a solicitor.
2. One of Melanie‟s students, Jason, came into her office and said he could not reach her this morning.
Melanie was confused because she had been in her office the whole time.
[D] She said: „Well, in that case, there must be something wrong with my telephone.‟
[I] She said that in that case, there must be something wrong with her telephone. [Q] Did Jason try to phone
Melanie in the morning?
3. A Blackwell book store has recently opened in Edinburgh. Today, Alexia, a young mother living nearby,
came in and asked for advice.
[D] She said: „It‟s my son‟s birthday tomorrow and I would like to purchase some storybooks on adventure.‟
[I] She said that it was her son‟s birthday soon and she would like to purchase some storybooks on adventure.
4. It was 5.30 pm and everybody was ready to leave the office. At one desk, Elaine was having a brief chat
with Steven about her work.
[D] She said: „Gosh, the amount of admin is killing me at the moment. I feel completely exhausted.‟
[I] She said that the amount of admin was killing her at the moment, and that she felt completely exhausted.
5. It was February 14. Carolyn and Tony were on a date at the newly opened Chinese restaurant. The past
couple of weeks, Carolyn was desperately trying to lose some weight.
[D] So she said to Tony: „Oh, I hope you don‟t mind if I‟m just having a starter.‟
[I] So she said to Tony that hopefully he wouldn‟t mind if she were just having a starter.
6. Ramona and Keith were postgraduate students in Linguistics. Today, they met in the local café and
Ramona started chatting about her favourite subjects.
[D] She said: „Latin is boring! I‟m much more interested in Eastern languages.‟
[I] She said that she found Latin boring and that she was much more interested in Eastern languages.
7. Derek‟s birthday was just a couple of days away and his girlfriend Ruby had arranged something special.
Tonight Ruby decided to unveil the plan that she kept hidden from him for so long.
[D] Over dinner, she said to Derek: „Well, your birthday is coming up soon, so I booked us tickets to London
to visit the opera.‟
[I] Over dinner, she revealed to Derek that since his birthday was coming up soon, she booked them tickets
to London to visit the opera.
[Q] Was Ruby‟s birthday coming up soon?
8. At the party on Friday night, some new faces were to be seen. Clare immediately caught Justin‟s eye,
which went over to her and offered her a drink.
[D] Clare replied: „Thanks very much, but I actually don‟t drink any alcohol.‟
[I] Clare thanked him and mentioned that she actually doesn‟t drink any alcohol.
[Q] Did Justin catch Clare‟s eye?
9. Alison and Nick travelled to Beijing during the Olympics. Alison was amazed by the fine fabric on offer
when they entered a big silk market.
[D] She said to Nick: „I could buy the whole lot of it! The silk feels so incredibly smooth.‟
[I] She said to Nick that she could buy the whole lot of it, and that the silk felt so incredibly smooth.
Page | 141
10. In the office, Ned told Daniela that his car had been damaged during an accident the other day. Daniela
was sorry to hear that and tried to comfort him.
[D] She said: „Well, at least you should get some money back from the insurance.‟
[I] She said that at least he should get some money back from the insurance.
[Q] Had Ned‟s car been damaged?
11. At the department store, Colleen was busy working. She was serving a customer who wanted to know
where he could buy some cosmetics for his wife.
[D] She replied: „Ah, cosmetics are actually on the second floor. Some of them are at a discount.‟
[I] She replied that cosmetics were actually on the second floor, adding that some of them were at a discount.
12. Jenny and Irvin were about to start a holiday trip to Barcelona. When they arrived at the airport to check
in, Jenny noticed that a heavy thunderstorm was brewing outside.
[D] She complained: „That‟s so unfair! I‟m sure the plane will be delayed.‟
[I] She complained that this was so unfair because she was sure that the plane would be delayed.
13. Britney is a student at the University of Glasgow. After a heavy snow in the afternoon, she was
complaining to her boyfriend James about the weather.
[D] She said: „I really hate the winter! It‟s always dark and the roads are too slippery.‟
[I] She said that she really hated the winter because it‟s always dark and the roads are too slippery.
14. Jasmine had been ill for a week so her brother Colin took her to the doctor. After an hour, Jasmine came
back to the waiting room looking rather disgruntled.
[D] She complained: „The treatment was absolutely useless! Next time, I‟d rather go to a specialist.‟
[I] She complained that the treatment was absolutely useless and that, next time, she‟d rather go to a
specialist.
15. A famous symphony was going to be played for free at the university concert hall. Kate and Andrew were
reading the advert, and Kate looked particularly interested.
[D] She said: „It sounds really good and it‟s for free! I‟m definitely going to that concert.‟
[I] She said that it sounded really good and that it was for free, adding that she would definitely go to that
concert.
16. On Valentine‟s Day, Simon at long last worked up the courage to propose to Emma. Emma wasn't sure,
but at the same time, she didn't want to upset him.
[D] So she explained: „You are a really nice bloke, Simon, but I‟m simply not ready for such a commitment.‟
[I] So she explained that although she thought he was a really nice bloke, she was simply not ready for such a
commitment.
17. Two college students were talking about on-line shopping. Fraser thought it was pointless and boring, but
Brenda couldn't disagree more.
[D] She said: „What!? On-line shopping is absolutely amazing! I just snapped up some really cheap DVDs on
Amazon.‟
[I] She said she found on-line shopping absolutely amazing, and that she had just snapped up some really
cheap DVDs on Amazon.
[Q] Did Fraser find on-line shopping boring?
18. It was quarter past seven and the Aberdeen Stock House was serving dinner. Debbie and Mike were very
pleased with the meal and Debbie in particular was totally smitten.
[D] She said: „We have to come here more often. The fish was really tasty and the soup was absolutely
delicious.‟
[I] She said that they should go there more often because the fish was really fresh and the soup was
absolutely delicious.
Page | 142
19. It was 6-year-old Herbert‟s first day at school. His mother Laura, who had never let him play with other
children before, was very anxious.
[D] She said to her husband: „I‟m really not sure whether Herbert is ready for this. I'd rather wait for another
year.‟
[I] She said to her husband that she was really not sure whether Herbert was ready for this, and that she'd
rather wait for another year.
20. In order to finish the project in time, Audrey had been working in the office from 7am to 8pm without
taking a break. Her colleague Sean just came back from dinner with a big smile on his face, which obviously
upset her a bit.
[D] She complained: „It‟s so unfair! My stomach has been rumbling all day. I could eat a whole elephant.‟
[I] She complained that it was so unfair because her stomach had been rumbling all day, and that she could
eat a whole elephant.
21. Cheryl and Barry were in the waiting lounge for their honeymoon flight to Paris. Shortly before boarding
time, Cheryl noticed that her handbag was damaged.
[D] She shouted: „Oh no! There's a hole in my bag! Don't tell me I've lost my passport!...‟
[I] She shouted that there was a hole in her bag. She was worried that she had lost her passport.
[Q] Was Barry‟s bag broken?
22. It was eight o‟ clock in the evening when the Britain‟s Got Talent live recording session had finished. On
their way out of the theatre, Sarah and Ross were having a giggle over some of the auditions.
[D] Sarah said: „The jugglers at the end were funny, weren‟t they? Quite pathetic, actually, but very
entertaining.‟
[I] Sarah found that the jugglers at the end were very funny – quite pathetic, actually, but very entertaining.
23. Jessica and John were enjoying a hot and sunny day at the holiday resort in Spain. When they were
heading towards their usual spot on the beach, Jessica was slightly overwhelmed by the heat.
[D] She said: „Phew, this is almost a bit too hot for me! Perhaps we should first go back for a drink under the
sunshade.‟
[I] She said that it was almost a bit too hot for her, and suggested to first go back for a drink under the
sunshade.
24. In the small town of Oban lies this charming little pub. One day, famous writer Aileen burst in with a big
smile on her face.
[D] She said: „Guess what – my new novel has been accepted for publication! I'll buy everyone a round.‟
[I] She said that her new novel had been accepted for publication and that she wanted to buy everyone a
round.
25. Theatrical agent Peter was visited by one of his clients. Shauna, a talented but struggling actress and
singer, was desperate for a part in the new musical.
[D] She said: „It‟s embarrassing to say, but without this job, I won't even be able to pay for electricity.‟
[I] She was embarrassed to say that without this job, she wouldn't even be able to pay for electricity.
[Q] Was it Mary who visited Peter?
26. It was Sunday afternoon when housewife Heather was tidying up the bedroom. After just two minutes,
she came out of the bedroom and confronted her husband Ben.
[D] She said: „I can‟t believe how messy you are! The first thing I found on the floor was your dirty pyjamas!‟
[I] She said that she could not believe how messy he was, and that the first thing she found on the floor was
his dirty pyjamas.
27. Morton needed to speak to his project supervisor Marianne at her office. When he asked whether she
could spare a minute, she seemed rather busy.
[D] She replied: „I‟m really sorry, but I have to finish marking some essays first.‟
[I] She replied that she was really sorry and that she had to finish marking some essays first.
Page | 143
28. It was Christmas Eve and everyone was drinking and dancing at the local pub. Judith was already quite
tipsy when she noticed that George was with a girl she had not seen before.
[D] She went over and complained to George: „You‟ve actually never told me that you had such a beautiful
sister.‟
[I] She went over and complained to George that he actually never told her that he had such a beautiful sister.
[Q] Was Judith George‟s sister?
29. Maureen had helped David a lot with his coursework, so David wanted to invite her for dinner. Asked
what kind of food she would fancy, Maureen first thought for a second.
[D] Then she replied: „I‟m normally not that adventurous, but perhaps I should try some sushi.‟
[I] Then she replied that although she would normally not be that adventurous, perhaps she should try some
sushi.
30. Jane was a new post-graduate at the department. When she signed up at the Sports Club, she was having a
conversation with Gregg about her favourite sports.
[D] She said: „Oh, I haven‟t practised a lot recently, but I used to be quite good at tennis.‟
[I] She said that although she had not practised a lot recently, she used to be quite good at tennis.
31. Leon and Betty were trying to get a mortgage for a new flat. While Betty was studying the relevant
newspaper adverts, Leon was browsing some web-sites.
[D] After a while, he said to Betty: „Hmm, these days, it looks as if nobody would give us an instant
cashback.‟
[I] After a while, he said to Betty that it looked as if nobody would give them an instant cashback.
32. Marcus and Helen were planning a trip to Europe for their wedding anniversary. Marcus was leaning
more towards Prague while Helen preferred Rome.
[D] Marcus explained: „Okay, Rome is nice, but the Czechs actually have much better beer.‟
[I] Marcus explained that Rome would be nice, but that the Czechs would actually have much better beer.
33. Bill and Dianne went out bowling together. When Dianne wanted to pick a ball of her favourite colour,
Bill looked a bit worried.
[D] He said: „That one looks quite heavy. I‟d rather pick the green one which I think is much smaller.‟
[I] He said that the ball looked quite heavy and that he‟d rather pick the green one which he thought was
much smaller.
[Q] Was Dianne going to pick a heavy ball at first?
34. Rob and Becky were buying duvet covers at the store. When Becky chose a bright red one, Rob was
completely against the colour.
[D] He said: „No way! Just imagine that every morning you'd be waking up in a sea of blood!‟
[I] He asked Becky just to imagine that every morning she would be waking up in a sea of blood.
[Q] Did Becky fancy a red duvet cover?
35. Harry and Leona were enjoying an afternoon walk in the park. When they were approaching the pond,
Harry was impressed by this peaceful place.
[D] He said: „This is so relaxing after a busy week! It is almost magical.‟
[I] He found that it was so relaxing after a busy week, and that it was almost magical.
36. Schoolteacher Edward recently caused a stir by marrying Lynda, one of his ex-pupils. Annoyed by all the
gossip, Edward told Lynda to simply not listen to what people say.
[D] He said: „I think there‟s nothing wrong about a teacher marrying a former student.‟
[I] He said that he thought there was nothing wrong about a teacher marrying a former student.
[Q] Had Lynda been Edward‟s teacher?
37. After making a small fortune in the lottery, Pamela bought quite an expensive car. Today in the car park,
her colleague Ralph was extremely impressed by her new possession.
[D] He said: „Wow! Is that your new car? It looks really stylish.‟
[I] He asked her whether that was her new car, adding that it looked really stylish.
[Q] Was Pamela impressed by Ralph‟s new car?
Page | 144
38. Brian invited some of his mates home and started making a mess in the kitchen. His mum was obviously
not happy about this, and Brian was quick to offer an excuse.
[D] He told her: „Well, we just had a couple of drinks because the pubs are already closed.‟
[I] He told her that they just had a couple of drinks because the pubs were already closed.
39. Daniel was injured in a car crash last week. At the hospital, he was quickly recovering. Today, he was
surprised that some distant relatives came to visit him.
[D] Not without sarcasm, Daniel said: „Oh, looks as if one has to get hurt to get some flowers.‟
[I] Not without sarcasm, Daniel said that it looked as if one had to get hurt to get some flowers.
40. Jacob had promised Cindy to buy a bonsai tree to decorate their living room. When Cindy wanted to
place the plant by the television, he tried to persuade her of an alternative option.
[D] He said: „Look, if you do that, there will be no sunlight for the bonsai, so it's better to place it by the
window.‟
[I] He said that if she did that, there would be no sunlight for the bonsai, so it was better to place it by the
window.
41. PhD student Ella was summoned to her supervisor Jim‟s office to give a report on her current progress.
When Ella asked for an extension, Jim looked concerned.
[D] He said: „Hmm, we really need those data in by next month for that conference.‟
[I] He said that they really needed those data in by next month for that conference.
[Q] Was Ella Jim‟s PhD student?
42. Albert was taking Joanne to the new Schwarzenegger movie. Albert was very excited, but since Joanne
seemed not very keen, he tried to persuade her to go with him.
[D] He said: „Come on! I‟m sure you will love it! The movie has some really cool special effects.‟
[I] He said he was sure she would love it, and that the movie had some really cool special effects.
43. Neil and Stephanie were visiting the local distillery. Neil was amazed by the range of different Whiskeys
produced, and he tried a very expensive Single Malt.
[D] He said: „Gee, that‟s a strong one! I‟m glad I won‟t have to drive us home.‟
[I] He said it was really strong and that he was glad he didn‟t have to drive them home.
[Q] Did Neil try a Single Malt?
44. Clive and his wife Molly visited the new seafood restaurant for the first time. Clive was extremely
anxious about food poisoning and did not want to have anything raw.
[D] So he said to Molly: „I don‟t care what other people might think, but I will order properly cooked oysters.‟
[I] So he said to Molly that, regardless of what other people might think, he would order properly cooked
oysters.
45. Luke and his friends were watching a movie at the cinema. Luke wasn‟t particularly keen on romantic
comedies, and he was complaining a lot after the film.
[D] He said: „God, that movie was terrible! I‟ve never been so bored in my life.‟
[I] He said that the movie was terrible and that he had never been so bored in his life.
46. On the train to London, Thomas was very nervous about his upcoming job interview. When he reached
for a cigarette to calm himself down, the middle-aged man next to him complained instantly.
[D] The man called out: „Oi! Smoking on public transport is illegal!‟
[I] He called out that smoking on public transport was illegal.
47. During a coffee break, university teachers Robert and Isabel met in the common room. Robert was
slightly irritated with the mess in the kitchen.
[D] He said to Isabel: „Look at this! Certain people must always leave their dirty mugs on the table.‟
[I] He said to Isabel that, obviously, certain people must always leave their dirty mugs on the table.
[Q] Was Isabel complaining about the dirty mugs?
Page | 145
48. School teachers Duncan and Liana were arguing about the curriculum. Apparently, Duncan was not a big
fan of Liana‟s modern style of teaching.
[D] He said: „If my own children were at this school, I‟d make sure they would never have to attend your
lessons.‟
[I] He said that if his own children were at this school, he‟d make sure they would never have to attend her
lessons.
49. It was getting quite late and most of Andrea‟s friends were about to leave her party. Scott in particular
seemed to have had a great time when he was leaving.
[D] He said: „Sorry about the crack in that glass. The party was absolutely fantastic.‟
[I] He apologised for the crack in that glass, adding that the party was absolutely fantastic.
[Q] Did Scott throw a party?
50. Doug always enjoyed a bit of gardening in his spare time. He recently bought a new lawnmower, but he
was quite frustrated because it simply refused to work.
[D] He complained: „Sod it! Things used to be much better in the past. All one can buy these days is rubbish!‟
[I] He complained that things used to be much better in the past, and that all one can buy these days was
rubbish.
51. After class, Roy had a serious conversation with one of his pupils. The boy had been accused of bullying
and Roy was giving him a caution.
[D] Roy said: „If the accusations are true, you will be in serious trouble. One cannot tolerate such kind of
behaviour.‟
[I] Roy said that if the accusations were true, the boy would be in serious trouble, and that one cannot
tolerate such kind of behaviour.
52. In the board room, assistant manager Craig was giving a presentation on how to improve the company‟s
sales figures. When he noticed that his boss was anything but impressed, Craig used his lack of experience as
an excuse.
[D] He said: „Sorry, this is actually my first ever presentation and I‟m probably not very persuasive.‟
[I] He said that this was actually his first ever presentation, and that he was probably not very persuasive.
53. Alice was working in a small antiques shop down the local high street. Today, a weird-looking man with
greasy hair and thick glasses came into the shop.
[D] He looked around and said: „You may be surprised to learn that I‟m a world-renowned collector of rare
pheasant paintings.‟
[I] He looked around and said that she might be surprised to learn that he was a world-renowned collector of
rare pheasant paintings.
54. Ray and Sandra were on a trip to the countryside for the weekend. Ray quite liked the outdoors and
particularly that rural smell.
[D] He said: „Aah, this is so much better than crowded cities and pollution.‟
[I] He said that it was so much better than crowded cities and pollution.
55. Earlier in the afternoon, Max and Janis were having a cigarette in front of the main entrance. Not
knowing what to say, Max started his typical conversation about the weather.
[D] He said: „Gosh, it‟s quite windy today – but at least it‟s not raining.‟
[I] He said that although it was quite windy today, at least it was not raining.
56. A journalist was interviewing Eric, the older brother of a famous pop diva. Eric did not really enjoy the
attention he was given and wasn‟t sure what to say.
[D] He proclaimed: „I actually don‟t see my sister very often these days. She is completely devoted to her
career.‟
[I] He proclaimed that he actually did not see his sister very often in those days, and that she was completely
devoted to her career.
Page | 146
57. Thomas went to the local McDonalds and incidentally met Sheena, who he knew from former Weight
Watchers meetings. Naturally, they were both quite embarrassed, but Thomas finally broke the silence.
[D] He said: „Weight Watchers is rubbish, isn‟t it. I‟ve tried so many diets but I‟m still overweight.‟
[I] He said that Weight Watchers was rubbish, and that he had tried so many diets but was still overweight.
58. A car-boot sale was taking place in the neighbourhood and Kenny was trying to get rid of some of the old
records he owned since the 1970s. A trendy young woman appeared to be very interested, so Kenny started
advertising.
[D] He said: „I can give you three albums for a fiver. You won‟t find any of that stuff on the internet.‟
[I] He said that he could give her three albums for a fiver, adding that she wouldn‟t find any of that stuff on
the internet.
59. A trendy night club had been recently opened near Marvin‟s flat and it proved instantly popular. However,
Marvin often felt disturbed by the noise, and this morning he was complaining to the neighbours about it.
[D] He said: „This is unbelievable! Every time I have to work early hours, they are having a bash!‟
[I] He said that every time he has to work early hours, they are having a bash.
60. Carl, a taxi driver, was about to drive home when suddenly he saw a man by the road frantically waving
his arms. Carl stopped and the man offered him double the fare if they make it to the airport within ten
minutes.
[D] Carl replied: „I‟ll try my best, mate, but cannot work any miracles.‟
[I] Carl replied that he would try his best, but that he could not work any miracles.
61. It was 11am in the morning when the fire alarm went off. Hearing people running down the corridors,
Mary grabbed her jacket and burst into Peter‟s office next door.
[D] She shouted: „Peter, quick, we have to leave immediately because the building is on fire!‟
[I] She urged Peter to leave immediately because the building was on fire.
62. It was early Tuesday morning when Claire was looking for the train tickets she left in one of her
handbags the other day. After searching for a couple of minutes she became worried and confused.
[D] She said to partner Gareth: „Erm, that‟s weird... I am pretty sure they were in that black leather bag, but
now they're gone.‟
[I] She said to her partner Gareth that it was weird because she was pretty sure that they were in that black
leather bag, but now they were gone.
63. Twenty-five year old Connie was going to sing in the quarter-final of a local talent competition. She was
extremely nervous before her performance, even though her mother had tried to calm her down.
[D] She said: „No! I can‟t do it! This is the end of the journey. I can‟t face the audience this time!‟
[I] She said that she could not do it and that was the end of the journey because she couldn‟t face the
audience this time.
64. At the beach, William and Phillip were enjoying the sunshine and a fresh sea breeze. Suddenly, William
noticed a piece of wood floating in the sea, with something lying on it that looked like a lifeless human body.
[D] Without hesitation, William exclaimed: „Phil, look! There's someone in trouble! We have to get some
help quickly!‟
[I] Without hesitation, William exclaimed that there was someone in trouble and that they had to get some
help quickly. [Q] Was it William who found the lifeless human body?
65. At Mr. Harris's house, Police constables Paul and Ken had some sad news to impart. Mrs. Harris's
husband, a well-known businessman, was killed in a car crash earlier this morning.
[D] When Paul spoke with her, he said: „Mrs. Harris, I am afraid that your husband was involved in a road
accident. He was announced dead at the scene.‟
[I] When Paul spoke with her, he said that her husband was involved in a road accident and that he was
announced dead at the scene.
Page | 147
66. On the train to London, young mother Olivia was taking her 5 year old daughter for a holiday. The little
girl could not understand why the youth who lit a cigarette at the station was scolded by other passengers.
[D] She asked: „If people don't like smoking, why don‟t they just hold their breath?‟
[I] She suggested that if people don't like smoking, they could just hold their breath.
[Q] Did the little girl think the youth could smoke at the station?
67. It was quarter past seven at the French restaurant, and head chef Pascal was extremely distressed. During
a routine check, the restaurant manager went berserk about the messy state of the kitchen.
[D] He said: „Is this how you want to make yourself a reputation?! It's a miracle that the health inspectors
haven't closed us down yet!‟
[I] He asked whether this would be the way he wanted to make himself a reputation, adding that it's a miracle
that the health inspectors haven't closed them down yet.
68. Teenagers Helen and Laura were shopping around at the summer sale. Helen was very fond of a black
skirt and shrieked when she found that it was heavily discounted.
[D] She called out: „Laura, look! This is exactly what I always wanted, and it's merely fifty quid!‟
[I] She called out to Laura that this is exactly what she always wanted, and that it was merely fifty quid.
69. On Sunday afternoon, the Riley‟s were going to the train station to pick up granddad who lived in the
neighbouring town. Twelve year-old Karen did not understand why they went so early if the train was
arriving at 6 pm.
[D] Mr Riley replied: „Well, just in case.. Remember last time when granddad forgot to set his watch
properly?‟
[I] Mr Riley replied that this was just in case, reminding her of the last time when granddad forgot to set his
watch properly.
70. A medical emergency tutorial was taking place in the operating theatre. The tutor was demonstrating the
steps to treat a heavy injury to the students.
[D] He explained: „When the patient comes in, first set up the life-support machine and then stop the
bleeding.‟
[I] He explained that when the patient comes in, they first should set up the life-support machine and then
stop the bleeding.
[Q] Is it correct to stop the bleeding first?
71. At Omaha Beach, allied soldiers braved artillery fire, mortar attacks and all-round danger. In response to
the enemy‟s gunfire, the captain of the 1st Division rallied the soldiers.
[D] He shouted: „Be brave, lads! Toss your grenades and open fire with your rifles! We must silence that
machine gun.‟
[I] He told the soldiers to toss their grenades and open fire with their rifles, for they must silence that
machine gun.
72. The West End festival was coming soon and the organisers were discussing details of the live television
broadcast. The current issue was where to set the television cameras and John was offering his plans.
[D] He said: „We might need television cameras along the High Street, and possibly even a helicopter to
cover the whole area.‟
[I] He suggested that they might need television cameras along the High Street and possibly even a helicopter
to cover the whole area.
73. Today in Northern Afghanistan, the United Nations Peacekeeping Force was ambushed by Taliban forces.
Although they returned fire immediately, the UN troops suffered heavy losses.
[D] The officer hastily called for reinforcements: „We are facing rocket and mortar fire in the region! We
urgently need armoured support!‟
[I] The officer hastily called for reinforcements, confirming that they were facing rocket and mortar fire in
the region and that they urgently needed armoured support.
74. It was midnight already and moonlight dripped through the leaves on the pavements. Robert and Chloe
just came back from a party, and Chloe was still full of energy.
[D] She patted Robert‟s cheek and said: „Stop yawning, my darling! The day is far from over yet.‟
[I] She patted Robert‟s cheek and said that he should stop yawning because the day was far from over yet.
Page | 148
75. Joseph, an extremely wealthy entrepreneur, was known to be addicted to horse racing. Tipped-off by an
insider, he placed a huge bet on one horse today, which failed to win the race, however.
[D] Joseph was now shouting furiously on the phone: „Where did your bloody information come from!? I've
lost nearly one million pounds!‟
[I] Joseph was now shouting furiously on the phone, asking where the information had come from, and
claiming that he had lost nearly one million pounds.
[Q] Was Joseph confident about the bet?
76. Due to the economic recession, Ron‟s company has lost a great quantity of international trades. His wife
noticed that every day he just slumped on the couch, with a tired and dull look on his face.
[D] She tried to comfort him by saying: „Why don't you stop your business for a while? We could spend
more time together, and there is certainly a positive side to that!‟
[I] She tried to comfort him by suggesting that if he stops his business for a while, they could have more time
together, and there was certainly a positive side to that.
77. At the airport, Carol and Fraser were boarding the flight to Barcelona for a conference. There were only 5
minutes left before the plane was taking off, and Carol started panicking because she couldn‟t find her
boarding card.
[D] She turned to Fraser and said: „Oh no! I can't find my stupid boarding pass! I must have left it at the dutyfree shop!‟
[I] She turned to Fraser and said that she couldn't find her boarding pass and that she must have left it at the
duty-free shop.
78. At Glasgow Royal Infirmary, family members were sitting around a bed, feeling sad. A very old man was
dying, and too weak to sit up. He wanted to say something, so his daughter placed a cushion under his head.
[D] He looked around slowly, thanked them for their coming, and said: „I am so happy to have the whole
family around me in my final hour.‟
[I] He looked around slowly, thanked them for their coming, and added that he was so happy to have the
whole family around him in his final hour.
79. Gerry and Herbert were attending a first aid tutorial where they were practising resuscitation procedures
using a plastic dummy. Gerry was busy remembering all the steps, but Herbert wouldn't bother and instead
explained his own strategy.
[D] He said: „Look, it's not necessary to go through all the steps. The important thing is to ensure that air
passages are not obstructed.‟
[I] He said that it's not necessary to go through all the steps, and that the important thing was to ensure that
air passages were not obstructed.
80. At House of Fraser, the manager was demonstrating her employees how to describe, in one sentence, the
special features of a product. She did it by performing a mock sales pitch in front of them.
[D] Holding up a jacket, she said: „This leather jacket is made of the finest materials and represents the latest
Italian couture.‟
[I] Holding up a jacket, she said that it was made of the finest materials and represented the latest Italian
couture.
[Q] Was the leather jacket from France?
81. A heavily armed robbery took place at the jewelleries‟ early this morning. The criminals escaped in a
blue van. Within five minutes, three police cars, an armed response unit, and an ambulance showed up at the
scene.
[D] The officer-in-charge ordered: „Every main road must be blocked immediately! And, for Christ's sake,
call the helicopter in!‟
[I] The officer-in-charge ordered that every main road must be blocked immediately, and that the helicopter
should be called in.
Page | 149
82. David was producing a TV documentary about the Vietnam War, and today, he was interviewing a 75
year-old Vietnam veteran. When David asked about the situation at the battle of Hue, the old soldier paused
for a while.
[D] Then, slowly, he said: „This battle was one of the longest and most brutal of the entire war. It's haunting
me in my dreams for the rest of my life.‟
[I] Then, slowly, he said that this battle was one of the longest and most brutal of the entire war, and that it
would be haunting him in his dreams for the rest of his life.
83. It was a Sunday in June, and a young couple were waiting for their relatives to arrive. Zoe was busy
making cucumber sandwiches and scones, when her husband noticed that her hands were actually shaking a
bit.
[D] She explained: „Don't worry, I'm OK! I'm just a little nervous because I really want everything to be
perfect.‟
[I] She explained that she was OK, and that she was just a little nervous because she really wanted
everything to be perfect.
84. University lecturer Bridget was responsible for admission to the English Literature classes. This morning,
she had to deal with yet another request from a student who wanted to change her course.
[D] The student said: „I‟m not sure whether Contemporary Women‟s Poetry is right for me. Is it possible to
switch to Sexual Identity in the Works of Oscar Wilde instead?‟
[I] The student said that she was not sure whether Contemporary Women‟s Poetry was right for her and
asked if it was possible to switch to Sexual Identity in the Works of Oscar Wilde instead.
85. In the computer shop, a customer wanted to get his laptop fixed because the Windows upgrade hadn‟t
installed correctly. The shop assistant had a quick look and noticed that the customer was using the Home
Edition.
[D] He said: „Well, I presume you've tried the wrong update. You can't really apply updates for XP
Professional to the Home Edition.‟
[I] He presumed that the customer had tried the wrong update, adding that one cannot really apply updates
for XP Professional to the Home Edition.
86. At Saint Jacob‟s hospital in Falmouth, Henry was waiting to have an X-ray taken. He fell off the bus this
morning and feared he broke his leg, but worst of all, he was also very anxious about X-rays.
[D] The doctor said: „No need to worry! We will give you a lead board to protect you from radiation.‟
[I] The doctor told him not to worry because they would give him a lead board to protect him from radiation.
87. The heat of the morning was intense but Charles and Lucy had managed to climb up a hill. From the top,
Charles could see right down to the beach on the other side.
[D] Excited, he said to Lucy: „Brilliant! Let's run down to the beach and take a plunge into the sea!‟
[I] Excited, he suggested to Lucy that they should run down to the beach and take a plunge into the sea. [Q]
Did Lucy suggest that they should take a plunge into the sea?
88. Mrs. Jones felt awkward to ask her son for help because she ran out of cash for the babysitter. She
pretended to enter her son‟s room 'by accident', trying to borrow £20.
[D] The boy replied: „You see, I'd love to help you out, but your credit history isn't what I would call
spotless.‟
[I] The boy replied that he would love to help her out, but that her credit history wasn't what he would call
spotless. [Q] Did Mrs. Jones successfully borrow the money?
89. Eleanor's son had always been a bit of a couch potato. Over dinner, Eleanor was passing a cutting from a
London newspaper to her husband.
[D] She said: „There, just read this article about child education! It exactly confirms my concerns about
watching too much TV.‟
[I] She asked him to read that article about child education, for it exactly confirmed her concerns about
watching too much TV.
Page | 150
90. The Smiths had recently suffered from some financial problems, so Mr. Smith was looking for a better
job. Today, he was going to have an interview with the manager of a big insurance company.
[D] Before he left, his wife said: „Best of luck, darling. Always remember that you are the best man for the
job!‟
[I] Before he left, his wife wished him best of luck, and advised him to remember that he is the best man for
the job.
Page | 151
Appendix 2 Materials for oral reading (Experiment 3)
[D] = direct speech; [I] = indirect speech; [VF] = voice focused; [VU] = voice unfocused
Item
1
2
3
Female reported speaker
Julie and Mark had been classmates and have not seen
each other for years.Today, they met in the local pub
and Julie started a conversation about career paths.
Male reported speaker
Mark and Julie had been classmates and have not seen
each other for years.Today, they met in the local pub
and Mark started a conversation about career paths.
[VF - D] She was very pleased with her job and
cheerfully talked about it, almost like a newly wedded
woman, saying: „My life has been amazing! After
merely three years, I‟m now a solicitor.‟
[VF - I] She was very pleased with her job and
cheerfully talked about it, almost like a newly wedded
woman, saying that her life had been amazing, and
that after merely three years, she now was a solicitor.
[VU - D] She was very pleased with her job and was
gesturing a lot, almost like a 12-year-old teenager,
saying: „My life has been amazing! After merely three
years, I‟m now a solicitor.‟
[VU - I] She was very pleased with her job and was
was gesturing a lot, almost like a 12-year-old teenager,
saying that her life had been amazing, and that after
merely three years, she now was a solicitor.
One of Melanie‟s students, Jason, came into her office
and said he could not reach her this morning. Melanie
could not believe it because she had been in her office
the whole time.
[VF - D] He was very pleased with his job and
cheerfully talked about it, almost like a newly-wed,
saying: „My life has been amazing! After merely three
years, I‟m now a solicitor.‟
[VF - I] He was very pleased with his job and
cheerfully talked about it, almost like a newly-wed,
saying that his life had been amazing, and that after
merely three years, he now was a solicitor.
[VU - D] He was very pleased with his job and was
gesturing a lot, almost like a 12-year-old teenager,
saying: „My life has been amazing! After merely three
years, I‟m now a solicitor.‟
[VU – I] He was very pleased with his job and was
was gesturing a lot, almost like a 12-year-old teenager,
saying that his life had been amazing, and that after
merely three years, he now was a solicitor.
One of Jason‟s students, Melanie, came into his office
and said she could not reach him this morning. Jason
could not believe it because he had been in his office
the whole time.
[VF – D] She replied with a sarcastic tone in her
voice: „Really?! Well, in that case, there must be
something wrong with my telephone.‟
[VF – I] She replied with a sarcastic tone in her voice,
saying that in that case, there must be something
wrong with her telephone.
[VU – D] She picked up the phone and listened
carefully, concluding: „Well, in that case, there must
be something wrong with my telephone.‟
[VU – I] She picked up the phone and listened
carefully, and concluded that in that case, there must
be something wrong with her telephone.
A Blackwell book store has recently opened in
Edinburgh.Today, Alexia, a young mother living
nearby, came in and looked for a present for her little
boy.
[VF – D] He replied with a sarcastic tone in his voice:
„Really?! Well, in that case, there must be something
wrong with my telephone.‟
[VF – I] He replied with a sarcastic tone in his voice,
saying that in that case, there must be something
wrong with his telephone.
[VU – D] He picked up the phone and listened
carefully, concluding: „Well, in that case, there must
be something wrong with my telephone.‟
[VU – I] He picked up the phone and listened
carefully, and concluded that in that case, there must
be something wrong with his telephone.
A Blackwell book store has recently opened in
Edinburgh.Today, Alex, a young father living nearby,
came in and looked for a present for his little boy.
[VF – D] Her voice was chirpy and melodic, saying:
„I'm very excited! It‟s my son‟s birthday tomorrow
and I want to surprise him with his favourite
storybooks on adventure.‟
[VF – I] Her voice was chirpy and melodic, saying
that she was very excited, for it was her son‟s birthday
tomorrow and she wanted to surprise him with his
favourite storybooks on adventure.
[VU – D] She said, with a sparkle in her eyes: „I'm
very excited! It‟s my son‟s birthday tomorrow and I
want to surprise him with his favourite storybooks on
adventure.‟
[VU – I] She said, with a sparkle in her eyes, that she
was very excited, for it was her son‟s birthday
tomorrow and she wanted to surprise him with his
favourite storybooks on adventure.
[VF – D] His voice was chirpy and melodic, saying:
„I'm very excited! It‟s my son‟s birthday tomorrow
and I want to surprise him with his favourite
storybooks on adventure.‟
[VF – I] His voice was chirpy and melodic, saying
that he was very excited, for it was his son‟s birthday
tomorrow and he wanted to surprise him with his
favourite storybooks on adventure.
[VU – D] He said, with a sparkle in his eyes: „I'm very
excited! It‟s my son‟s birthday tomorrow and I want to
surprise him with his favourite storybooks on
adventure.‟
[VU – I] He said, with a sparkle in his eyes, that he
was very excited, for it was his son‟s birthday
tomorrow and he wanted to surprise him with his
favourite storybooks on adventure.
Page | 152
4
5
6
It was 5.30 pm and everybody was ready to leave the
office.At one desk, Elaine was having a brief chat with
Steven about her work.
It was 5.30 pm and everybody was ready to leave the
office.At one desk, Steven was having a brief chat
with Elaine about his work.
[VF – D] She complained, with a hint of reluctance in
her voice: „Gosh, the amount of admin is killing me at
the moment. I feel completely exhausted!‟
[VF – I] She complained, with a hint of reluctance in
her voice, that the amount of admin was killing her at
the moment, and that she felt completely exhausted.
[VU – D] She complained, with her face looking pale
and tired: „Gosh, the amount of admin is killing me at
the moment. I feel completely exhausted.‟
[VU – I] She complained, with her face looking pale
and tired, that the amount of admin was killing her at
the moment, and that she felt completely exhausted.
It was February 14. Carolyn and Tony were on a date
at the newly opened Chinese restaurant. The past
couple of weeks, Carolyn was desperately trying to
lose some weight.
[VF – D] He complained, with a hint of reluctance in
his voice: „Gosh, the amount of admin is killing me at
the moment. I feel completely exhausted!‟
[VF – I] He complained, with a hint of reluctance in
his voice, that the amount of admin was killing him at
the moment, and that he felt completely exhausted.
[VU – D] He complained, with his face looking pale
and tired: „Gosh, the amount of admin is killing me at
the moment. I feel completely exhausted.‟
[VU – I] He complained, with his face looking pale
and tired, that the amount of admin was killing him at
the moment, and that he felt completely exhausted.
It was February 14. Tony and Caroline were on a date
at the newly opened Chinese restaurant. The past
couple of weeks, Tony was desperately trying to lose
some weight.
[VF – D] She didn't want to disappoint Tony, so she
said to him, softly and gently: „Oh darling, I hope you
don‟t mind if I‟m just having a starter.‟
[VF – I] She didn't want to disappoint Tony, so she
said to him, softly and gently, that hopefully he
wouldn‟t mind if she was just having a starter.
[VU – D] She didn't want to disappoint Tony, so she
said to him, with an apologetic smile: „Oh darling, I
hope you don‟t mind if I‟m just having a starter.‟
[VU – I] She didn't want to disappoint Tony, so she
said to him, with an apologetic smile, that hopefully he
wouldn‟t mind if she was just having a starter.
Ramona and Keith were postgraduate students in
Linguistics. Today, they met in the local café and
Ramona started chatting about their favourite subjects.
[VF – D] He didn't want to disappoint Caroline, so he
said to her, softly and gently: „Oh darling, I hope you
don‟t mind if I‟m just having a starter.‟
[VF – I] He didn't want to disappoint Caroline, so he
said to her, softly and gently, that hopefully she
wouldn‟t mind if he was just having a starter.
[VU – D] He didn't want to disappoint Caroline, so he
said to her, with an apologetic smile: „Oh darling, I
hope you don‟t mind if I‟m just having a starter.‟
[VU – I] He didn't want to disappoint Caroline, so he
said to her, with an apologetic smile, that hopefully
she wouldn‟t mind if he was just having a starter.
Keith and Ramona were postgraduate students in
Linguistics. Today, they met in the local café and
Keith started chatting about their favourite subjects.
[VF – D] When the topic switched to Latin, she
sounded really upset with a moaning voice: „Latin is
so boring! I‟m much more interested in Eastern
languages.‟
[VF – I] When the topic switched to Latin, she
sounded really upset with a moaning voice, saying that
she found Latin so boring and that she was much more
interested in Eastern languages.
[VU – D] When the topic switched to Latin, she had a
really upset look on her face: „Latin is so boring! I‟m
much more interested in Eastern languages.‟
[VU – I] When the topic switched to Latin, she had a
really upset look on her face, saying that she found
Latin so boring and that she was much more interested
in Eastern languages.
[VF – D] When the topic switched to Latin, he
sounded really upset with a moaning voice: „Latin is
so boring! I‟m much more interested in Eastern
languages.‟
[VF – I] When the topic switched to Latin, he sounded
really upset with a moaning voice, saying that he
found Latin so boring and that he was much more
interested in Eastern languages.
[VU – D] When the topic switched to Latin, he had a
really upset look on his face: „Latin is so boring! I‟m
much more interested in Eastern languages.‟
[VU – I] When the topic switched to Latin, he had a
really upset look on his face, saying that he found
Latin so boring and that he was much more interested
in Eastern languages.
Page | 153
7
8
9
Derek‟s birthday was just a couple of days away and
his girlfriend Ruby had arranged something special.
Tonight Ruby decided to unveil the plan that she kept
hidden from him for so long.
Ruby‟s birthday was just a couple of days away and
her boyfriend Derek had arranged something special.
Tonight Derek decided to unveil the plan that he kept
hidden from her for so long.
[VF- D] Over dinner, she drew Derek's attention with
a chirpy tone in her voice: „Look, darling, your
birthday is coming up soon, so... I booked us tickets to
London to visit the opera!‟
[VF – I] Over dinner, she drew Derek's attention with
a chirpy tone in her voice, saying that since his
birthday was coming up soon, she booked them tickets
to London to visit the opera.
[VU – D] Over dinner, she drew Derek's attention with
luring gesture: „Look, darling, your birthday is coming
up soon, so... I booked us tickets to London to visit the
opera.‟
[VU – I] Over dinner, she drew Derek's attention with
a luring gesture, saying that since his birthday was
coming up soon, she booked them tickets to London to
visit the opera.
At the party on Friday night, some new faces were to
be seen. Clare immediately caught Justin‟s eye, who
went over to her and offered her a drink.
[VF – D] Over dinner, he drew Ruby's attention with a
chirpy tone in his voice: „Look, darling, your birthday
is coming up soon, so... I booked us tickets to London
to visit the opera!‟
[VF – I] Over dinner, he drew Ruby's attention with a
chirpy tone in his voice, saying that since her birthday
was coming up soon, he booked them tickets to
London to visit the opera.
[VU – D] Over dinner, he beckoned Ruby over:
„Look, darling, your birthday is coming up soon, so... I
booked us tickets to London to visit the opera.‟
[VF – D] „Oh?!‟ she sounded a little flatterred:
„Thanks very much, but I actually don‟t drink any
alcohol.‟
[VF – I] Clare sounded a little flattered. She thanked
him and mentioned that she actually doesn‟t drink any
alcohol.
[VU – D] „Oh?!‟ Clare smiled like a blooming
sunflower, „Thanks very much, but I actually don‟t
drink any alcohol.‟
[VU – I] Clare smiled like a blooming sunflower. She
thanked him and mentioned that she actually doesn‟t
drink any alcohol.
Alison and Nick travelled to Beijing during the
Olympics. Alison was amazed by the fine fabric on
offer when they entered a big silk market.
[VF – D] „Oh?!‟ he sounded a little flattered: „Thanks
very much, but I actually don‟t drink any alcohol.‟
[VF – D] Her voice was vibrating frantically: „Oh my
goodness! I could buy the whole lot of it! The silk
feels so incredibly smooth.‟
[VF – I] Her voice was vibrating frantically, saying
that she could buy the whole lot of it, and that the silk
felt so incredibly smooth.
[VU – D] Her eyes were wide open, like two plump
strawberries: „Oh my goodness! I could buy the whole
lot of it! The silk feels so incredibly smooth.‟
[VU – I] Her eyes were wide open, like two plump
strawberries when she said that she could buy the
whole lot of it, and that the silk felt so incredibly
smooth.
[VF – D] His voice was vibrating frantically: „Oh my
goodness! I could buy the whole lot of it! The silk
feels so incredibly smooth.‟
[VF – I] His voice was vibrating frantically, saying
that he could buy the whole lot of it, and that the silk
felt so incredibly smooth.
[VU – D] His eyes were wide open, like two plump
strawberries: „Oh my goodness! I could buy the whole
lot of it! The silk feels so incredibly smooth.‟
[VU – I] His eyes were wide open, like two plump
strawberries when he said that he could buy the whole
lot of it, and that the silk felt so incredibly smooth.
[VU – I] Over dinner, he beckoned Ruby over, saying
that since his birthday was coming up soon, she
booked them tickets to London to visit the opera.
At the party on Friday night, some new faces were to
be seen. Justin immediately caught Claire‟s eye, who
went over to him and offered him a drink.
[VF – I] Justin sounded a little flattered. He thanked
her and mentioned that he actually doesn‟t drink any
alcohol.
[VU – D] „Oh?!‟ Justin smiled politely, „Thanks very
much, but I actually don‟t drink any alcohol.‟
[VU – I] Justin smiled politely. He thanked her and
mentioned that he actually doesn‟t drink any alcohol.
Nick and Alison travelled to Beijing during the
Olympics. Nick was amazed by the fine fabric on offer
when they entered a big silk market.
Page | 154
10
11
12
In the office, Ned told Daniela that his car had been
damaged during an accident the other day. Daniela
was sorry to hear that and tried to comfort him.
In the office, Daniela told Ned that her car had been
damaged during an accident the other day. Ned was
sorry to hear that and tried to comfort her.
[VF –D] She said in a calming and reassuring tone:
„Well, at least you should get some money back from
the insurance.‟
[VF – I] She said in a calming and reassuring tone:
that at least he should get some money back from the
insurance.
[VU – D] She said, with her hand gently patting his
back: „Well, at least you should get some money back
from the insurance.‟
[VU – I] She said, with her hand gently patting his
back, that at least he should get some money back
from the insurance.
At the department store, Colleen was busy working.
She was serving a customer who wanted to know
where he could buy some cosmetics for his wife.
[VF – D] He said in a calming and reassuring tone:
„Well, at least you should get some money back from
the insurance.‟
[VF – I] He said in a calming and reassuring tone: that
at least she should get some money back from the
insurance.
[VU – D] He said, with his hand gently patting her
back: „Well, at least you should get some money back
from the insurance.‟
[VU – I] He said, with his hand gently patting her
back, that at least she should get some money back
from the insurance.
At the department store, Colin was busy working. He
was serving a customer who wanted to know where
she could buy some cosmetics for her friend.
[VF – D] She replied in a polite and clear voice: „Ah,
cosmetics are actually on the second floor. Some of
them are at a discount.‟
[VF – I] She replied in a polite and clear voice that
cosmetics were actually on the second floor, adding
that some of them were at a discount.
[VU – D] She replied with a professional smile: „Ah,
cosmetics are actually on the second floor. Some of
them are at a discount.‟
[VU – I] She replied with a professional smile that
cosmetics were actually on the second floor, adding
that some of them were at a discount.
Jenny and Irvin were about to start a holiday trip to
Barcelona. When they arrived at the airport to check
in, Jenny noticed that a heavy thunderstorm was
brewing outside.
[VF – D] He replied in a polite and clear voice: „Ah,
cosmetics are actually on the second floor. Some of
them are at a discount.‟
[VF – I] He replied in a polite and clear voice that
cosmetics were actually on the second floor, adding
that some of them were at a discount.
[VU – D] He replied with a professional smile: „Ah,
cosmetics are actually on the second floor. Some of
them are at a discount.‟
[VU – I] He replied with a professional smile that
cosmetics were actually on the second floor, adding
that some of them were at a discount.
Irvin and Jenny were about to start a holiday trip to
Barcelona. When they arrived at the airport to check
in, Irvin noticed that a heavy thunderstorm was
brewing outside.
[VF – D] She became almost hysterical, squeaking:
„That‟s so unfair! I‟m sure the plane will be delayed.‟
[VF – I] She became almost hysterical, squeaking that
this was so unfair because she was sure that the plane
would be delayed.
[VU – D] She complained with a very disappointed
look on her face: „That‟s so unfair! I‟m sure the plane
will be delayed.‟
[VU – I ] She complained, with a very disappointed
look on her face, that this was so unfair because she
was sure that the plane would be delayed.
[VF – D] He became almost hysterical, growling:
„That‟s so unfair! I‟m sure the plane will be delayed.‟
[VF – I] He became almost hysterical, growling that
this was so unfair because he was sure that the plane
would be delayed.
[VU – D] He complained with a very disappointed
look on his face: „That‟s so unfair! I‟m sure the plane
will be delayed.‟
[VU – I] He complained, with a very disappointed
look on his face, that this was so unfair because he was
sure that the plane would be delayed.
Page | 155
13
14
15
Britney is a student at the University of Glasgow.
After a heavy snow in the afternoon, she was
complaining to her boyfriend James about the weather
on their way home.
James is a student at the University of Glasgow. After
a heavy snow in the afternoon, he was complaining to
his girlfriend Britney about the weather on their way
home.
[VF – D] Her voice sounded very grumpy and
unpleasant: „I really hate the winter! It‟s always dark
and the roads are too slippery.‟
[VF – I] Her voice sounded very grumpy and
unpleasant and unpleasant, saying that she really hated
the winter because it‟s always dark and the roads are
too slippery.
[VU – D] She said, with an unpleasant look on her
face: „I really hate the winter! It‟s always dark and the
roads are too slippery.‟
[VU – I] She said, with an unpleasant look on her
face, that she really hated the winter because it‟s
always dark and the roads are too slippery.
Jasmine had been ill for a week so her brother Colin
took her to the doctor. After an hour, Jasmine came
back to the waiting room looking rather disgruntled.
[VF – D] His voice sounded very grumpy and
unpleasant: „I really hate the winter! It‟s always dark
and the roads are too slippery.‟
[VF – I] His voice sounded very grumpy and
unpleasant, saying that he really hated the winter
because it‟s always dark and the roads are too slippery.
[VU – D] He said, with an unpleasant look on his face:
„I really hate the winter! It‟s always dark and the roads
are too slippery.‟
[VU – I] He said, with an unpleasant look on his face,
that he really hated the winter because it‟s always dark
and the roads are too slippery.
[VF – D] She complained in a sharp and irate voice:
„The treatment was absolutely useless! Next time, I‟d
rather go to a specialist.‟
[VF – I] She complained in a sharp and irate voice
that the treatment was absolutely useless and that, next
time, she‟d rather go to a specialist.
[VU – D] She complained with an annoyed look on
her face: „The treatment was absolutely useless! Next
time, I‟d rather go to a specialist.‟
[VU – I] She complained, with an annoyed look on
her face, that the treatment was absolutely useless and
that, next time, she‟d rather go to a specialist.
A famous symphony was going to be played for free at
the university concert hall. Kate and Andrew were
reading the advert, and Kate looked particularly
interested.
[VF – D] He complained in a sharp and irate voice:
„The treatment was absolutely useless! Next time, I‟d
rather go to a specialist.‟
[VF – I] He complained in a sharp and irate voice that
the treatment was absolutely useless and that, next
time, he‟d rather go to a specialist.
[VU – D] He complained with an annoyed look on his
face: „The treatment was absolutely useless! Next
time, I‟d rather go to a specialist.‟
[VU – I] He complained, with an annoyed look on his
face, that the treatment was absolutely useless and
that, next time, he‟d rather go to a specialist.
A famous symphony was going to be played for free at
the university concert hall. Andrew and Kate were
reading the advert, and Andrew looked particularly
interested.
[VF – D] Her voice went cheerful and excited: „It
sounds really good and it‟s for free! I‟m definitely
going to that concert.‟
[VF – I] Her voice went cheerful and excited, saying
that it sounded really good and that it was for free,
adding that she would definitely go to that concert.
[VU – D] With her eyes wide open, she said: „It
sounds really good and it‟s for free! I‟m definitely
going to that concert.‟
[VU – I] With her eyes wide open, she said that it
sounded really good and that it was for free, adding
that she would definitely go to that concert.
[VF – D] His voice went cheerful and excited: „It
sounds really good and it‟s for free! I‟m definitely
going to that concert.‟
[VF – I] His voice went cheerful and excited, saying
that it sounded really good and that it was for free,
adding that he would definitely go to that concert.
[VU – D] With his eyes wide open, he said: „It sounds
really good and it‟s for free! I‟m definitely going to
that concert.‟
[VU – I] With his eyes wide open, he said that it
sounded really good and that it was for free, adding
that he would definitely go to that concert.
Sam had been ill for a week so his sister Jasmine took
him to the doctor. After an hour, Sam came back to the
waiting room looking rather disgruntled.
Page | 156
16
17
18
On Valentine‟s Day, Simon at long last worked up the
courage to propose to Emma. Emma wasn't sure, but at
the same time, she didn't want to upset him.
On Valentine‟s Day, Emma at long last worked up the
courage to express her affection for Simon. Simon
wasn't sure, but at the same time, he didn't want to
upset her.
[VF – D] So she explained with a soft and gentle
voice: „You are a really nice bloke, Simon, but I‟m
simply not ready for such a commitment.‟
[VF – I] So she explained with a soft and gentle voice
that although she thought he was a really nice bloke,
she was simply not ready for such a commitment.
[VU – D] So she returned an apologetic smile,
explaining: „You are a really nice bloke, Simon, but
I‟m simply not ready for such a commitment.‟
[VU – I] So she returned an apologetic smile,
explaining that although she thought he was a really
nice bloke, she was simply not ready for such a
commitment.
Two college students were talking about on-line
shopping. Fraser thought it was pointless and boring,
but Brenda couldn't disagree more.
[VF – D] So he explained with a soft and gentle voice:
„You are a really nice girl, Emma, but I‟m simply not
ready for a relationship.‟
[VF – I] So he explained with a soft and gentle voice
that although he thought she was a really nice girl, he
was simply not ready for a relationship.
[VU – D] So he returned an apologetic smile,
explaining: „You are a really nice girl, Emma, but I‟m
simply not ready for a relationship.‟
[VU – I] So he returned an apologetic smile,
explaining that although he thought she was a really
nice girl, he was simply not ready for a relationship.
[VF – D] She raised her voice in disagreement:
„What!? On-line shopping is absolutely amazing! I just
snapped up some really cheap DVDs on Amazon.‟
[VF – I] She raised her voice in disagreement, saying
that she found on-line shopping absolutely amazing,
and that she had just snapped up some really cheap
DVDs on Amazon.
[VU – D] She said, gesturing frantically: „What!? Online shopping is absolutely amazing! I just snapped up
some really cheap DVDs on Amazon.‟
[VU – I] She said, gesturing frantically, that she found
on-line shopping absolutely amazing, and that she had
just snapped up some really cheap DVDs on Amazon.
[VF – D] He raised his voice in disagreement: „What!?
On-line shopping is absolutely amazing! I just snapped
up some really cheap DVDs on Amazon.‟
[VF – I] He raised his voice in disagreement, saying
that he found on-line shopping absolutely amazing,
and that he had just snapped up some really cheap
DVDs on Amazon.
[VU – D] He said, gesturing frantically: „What!? Online shopping is absolutely amazing! I just snapped up
some really cheap DVDs on Amazon.‟
[VU – I] He said, gesturing frantically, that he found
on-line shopping absolutely amazing, and that he had
just snapped up some really cheap DVDs on Amazon.
It was quarter past seven and the Aberdeen Stock
House was serving dinner. Debbie and Mike were very
pleased with the meal and Debbie in particular was
totally smitten.
It was quarter past seven and the Aberdeen Stock
House was serving dinner. Mike and Debbie were very
pleased with the meal and Mike in particular was
totally smitten.
[VF – D] Her voice was honeyed and passionate:
„Darling, we have to come here more often. The fish
was really tasty and the soup was absolutely
delicious.‟
[VF – I] Her voice was honeyed and passionate,
saying that they should go there more often because
the fish was really fresh and the soup was absolutely
delicious.
[VU – D] There was a pleased look on her face while
she said: „Darling, we have to come here more often.
The fish was really tasty and the soup was absolutely
delicious.‟
[VU – I] There was a pleased look on her face while
she said that they should go there more often because
the fish was really fresh and the soup was absolutely
delicious.
[VF – D] His voice was full of delight: „Darling, we
have to come here more often. The fish was really
tasty and the soup was absolutely delicious.‟
Two college students were talking about on-line
shopping. Brenda thought it was pointless and boring,
but Fraser couldn't disagree more.
[VF – I] His voice was full of delight, saying that they
should go there more often because the fish was really
fresh and the soup was absolutely delicious.
[VU – D] There was a pleased look on his face while
he said: „Darling, we have to come here more often.
The fish was really tasty and the soup was absolutely
delicious.‟
[VU – I] There was a pleased look on his face while
he said that they should go there more often because
the fish was really fresh and the soup was absolutely
delicious.
Page | 157
19
20
21
It was 6-year-old Herbert‟s first day at school. His
mother Laura, who had never let him play with other
children before, was very anxious.
It was 6-year-old Laura‟s first day at school. His father
Herbert, who had never let her play with other children
before, was very anxious.
[VF – D] She was talking to her husband with an
apprehensive tone in her voice: „I‟m really not sure
whether Herbert is ready for this. I'd rather wait for
another year.‟
[VF – I] She was talking to her husband, with an
apprehensive tone in her voice, saying that she was
really not sure whether Herbert was ready for this, and
that she'd rather wait for another year.
[VU – D] She said to her husband, with an
apprehensive look on her face: „I‟m really not sure
whether Herbert is ready for this. I'd rather wait for
another year.‟
[VU – I] She said to her husband, with an
apprehensive look on her face, that she was really not
sure whether Herbert was ready for this, and that she'd
rather wait for another year.
In order to finish the project in time, Audrey had been
working in the office from 7am to 8pm without taking
a break. Her colleague Sean just came back from
dinner with a big smile on his face, which obviously
upset her a bit.
[VF – D] He was talking to his wife with an
apprehensive tone in his voice: „I‟m really not sure
whether Laura is ready for this. I'd rather wait for
another year.‟
[VF – I] He was talking to his wife, with an
apprehensive tone in his voice, saying that he was
really not sure whether Laura was ready for this, and
that he'd rather wait for another year.
[VU – D] He said to his wife, with an apprehensive
look on his face: „I‟m really not sure whether Laura is
ready for this. I'd rather wait for another year.‟
[VF – D] She complained, with her voice turned
screechy: „It‟s so unfair! My stomach had been
rumbling all day. I could eat a whole elephant.‟
[VF – I] She complained, with her voice turned
screechy, that it was so unfair because her stomach
had been rumbling all day, and that she could eat a
whole elephant.
[VU – D] She complained, with an expression of
discontent: „It‟s so unfair! My stomach had been
rumbling all day. I could eat a whole elephant.‟
[VU – I] She complained, with an expression of
discontent, that it was so unfair because her stomach
had been rumbling all day, and that she could eat a
whole elephant.
Cheryl and Barry were in the waiting lounge for their
honeymoon flight to Paris. Shortly before boarding
time, Cheryl noticed that her handbag was damaged.
[VF – D] He complained, with his voice turned
screechy: „It‟s so unfair! My stomach had been
rumbling all day. I could eat a whole elephant.‟
[VF – I] He complained, with his voice turned
screechy, that it was so unfair because his stomach had
been rumbling all day, and that he could eat a whole
elephant.
[VU – D] He complained, with an expression of
discontent: „It‟s so unfair! My stomach had been
rumbling all day. I could eat a whole elephant.‟
[VU – I] He complained, with an expression of
discontent, that it was so unfair because his stomach
had been rumbling all day, and that he could eat a
whole elephant.
Barry and Cheryl were in the waiting lounge for their
honeymoon flight to Paris. Shortly before boarding
time, Barry noticed that his briefcase was damaged.
[VF – D] She was shocked and bursted into a
panicking shout: „Oh no! There's a hole in my bag!
Don't tell me I've lost my passport!...‟
[VF – I] She was shocked and bursted into a
panicking shout that there was a hole in her bag. She
was worried that she had lost her passport.
[VU –D] She was grabbing her hair, obviously
shocked: „Oh no! There's a hole in my bag! Don't tell
me I've lost my passport!...‟
[VU – I] She was grabbing her hair, obviously
shocked, saying that there was a hole in her bag. She
was worried that she had lost her passport.
[VF – D] He was shocked and bursted into a panicking
shout: „Oh no! There's a hole in my briefcase! Don't
tell me I've lost my passport!...‟
[VF – I] He was shocked and bursted into a panicking
shout that there was a hole in his briefcase. He was
worried that he had lost his passport.
[VU – D] He was grabbing his hair, obviously
shocked: „Oh no! There's a hole in my briefcase! Don't
tell me I've lost my passport!...‟
[VU – I] He was grabbing his hair, obviously shocked,
saying that there was a hole in his briefcase. He was
worried that he had lost his passport.
[VU – I] He said to his wife, with an apprehensive
look on his face, that he was really not sure whether
Laura was ready for this, and that he'd rather wait for
another year.
In order to finish the project in time, Sean had been
working in the office from 7am to 8pm without taking
a break. His colleague Audrey just came back from
dinner with a big smile on her face, which obviously
upset him a bit.
Page | 158
22
23
24
It was eight o‟ clock in the evening when the Britain‟s
Got Talent live recording session had finished. On
their way out of the theatre, Sarah and Ross were
having a giggle over some of the auditions.
It was eight o‟ clock in the evening when the Britain‟s
Got Talent live recording session had finished. On
their way out of the theatre, Ross and Sarah were
having a giggle over some of the auditions.
[VF – D] Still chuckling, Sarah said: „The jugglers at
the end were funny, weren‟t they? Quite pathetic,
actually, but very entertaining.‟
[VF –I] Still chuckling, Sarah said that the jugglers at
the end were very funny – quite pathetic, actually, but
very entertaining.
[VU – D] Sarah said, with a bemused look on her face:
„The jugglers at the end were funny, weren‟t they?
Quite pathetic, actually, but very entertaining.‟
[VU – I] Sarah said, with a bemused look on her face,
that the jugglers at the end were very funny – quite
pathetic, actually, but very entertaining.
Jessica and John were enjoying a hot and sunny day at
the holiday resort in Spain. When they were heading
towards their usual spot on the beach, Jessica was
slightly overwhelmed by the heat.
[VF – D] Still chuckling, Ross said: „The jugglers at
the end were funny, weren‟t they? Quite pathetic,
actually, but very entertaining.‟
[VF – I] Still chuckling, Ross said that the jugglers at
the end were very funny – quite pathetic, actually, but
very entertaining.
[VU – D] Ross said, with a bemused look on his face:
„The jugglers at the end were funny, weren‟t they?
Quite pathetic, actually, but very entertaining.‟
[VU – I] Ross said, with a bemused look on his face,
that the jugglers at the end were very funny – quite
pathetic, actually, but very entertaining.
John and Jessica were enjoying a hot and sunny day at
the holiday resort in Spain.When they were heading
towards their usual spot on the beach, John was
slightly overwhelmed by the heat.
[VF – D] She sounded exhausted when she said:
„Phew, this is almost a bit too hot for me! Perhaps we
should first go back for a drink under the sunshade.‟
[VF – I] She sounded exhausted when she said that it
was almost a bit too hot for her, and suggested to first
go back for a drink under the sunshade.
[VU – D] She said, wiping her forehead with a towel:
„Phew, this is almost a bit too hot for me! Perhaps we
should first go back for a drink under the sunshade.‟
[VU – I] She said, wiping her forehead with a towel,
that it was almost a bit too hot for her, and suggested
to first go back for a drink under the sunshade.
In the small town of Oban lies this charming little pub.
One day, famous writer Aileen burst in with a big
smile on her face.
[VF – D] He sounded exhausted when he said: „Phew,
this is almost a bit too hot for me! Perhaps we should
first go back for a drink under the sunshade.‟
[VF – I] He sounded exhausted when he said that it
was almost a bit too hot for him, and suggested to first
go back for a drink under the sunshade.
[VU – D] He said, wiping his forehead with a towel:
„Phew, this is almost a bit too hot for me! Perhaps we
should first go back for a drink under the sunshade.‟
[VU – I] He said, wiping his forehead with a towel,
that it was almost a bit too hot for him, and suggested
to first go back for a drink under the sunshade.
In the small town of Oban lies this charming little
pub.One day, famous writer Alan burst in with a big
smile on his face.
[VF – D] Her voice burst with joy: „Guess what – my
new novel has been accepted for publication! I'll buy
everyone a round.‟
[VF – I] Her voice burst with joy, saying that her new
novel had been accepted for publication and that she
wanted to buy everyone a round.
[VU – D] She said, waving her arms: „Guess what –
my new novel has been accepted for publication! I'll
buy everyone a round.‟
[VU – I] She said, waving her arms, that her new
novel had been accepted for publication and that she
wanted to buy everyone a round.
[VF – D] His voice burst with joy: „Guess what – my
new novel has been accepted for publication! I'll buy
everyone a round.‟
[VF – I] His voice burst with joy, saying that his new
novel had been accepted for publication and that he
wanted to buy everyone a round.
[VU – D] He said, waving his arms: „Guess what – my
new novel has been accepted for publication! I'll buy
everyone a round.‟
[VU – I] He said, waving his arms, that his new novel
had been accepted for publication and that he wanted
to buy everyone a round.
Page | 159
25
26
27
Theatrical agent Peter was visited by one of his clients.
Shauna, a talented but struggling actress and singer,
was desperate for a part in the new musical.
[VF – D] She explained, with a hint of anxiety in her
voice: „It‟s embarrassing to say, but without this job, I
won't even be able to pay for electricity.‟
[VF – I] She explained, with a hint of anxiety in her
voice, that it was embarrassing to say that without this
job, she wouldn't even be able to pay for electricity.
[VU – D] She explained, with her head swivelled
round to avoid eye contact: „It‟s embarrassing to say,
but without this job, I won't even be able to pay for
electricity.‟
[VU – I] She explained, with her head swivelled round
to avoid eye contact, that it was embarrassing to say
that without this job, she wouldn't even be able to pay
for electricity.
It was Sunday afternoon when housewife Heather was
tidying up the bedroom. After just two minutes, she
came out of the bedroom and confronted her husband
Ben.
Theatrical agent Shauna was visited by one of her
clients. Peter, a talented but struggling young actor and
singer, was desperate for a part in the new musical.
[VF – D] He explained, with a hint of anxiety in his
voice: „It‟s embarrassing to say, but without this job, I
won't even be able to pay for electricity.‟
[VF – I] He explained, with a hint of anxiety in his
voice, that it was embarrassing to say that without this
job, he wouldn't even be able to pay for electricity.
[VU – D] He explained, with his head swivelled round
to avoid eye contact: „It‟s embarrassing to say, but
without this job, I won't even be able to pay for
electricity.‟
[VU – I] He explained, with his head swivelled round
to avoid eye contact, that it was embarrassing to say
that without this job, he wouldn't even be able to pay
for electricity.
It was Sunday afternoon when father Ben was tidying
up his daughter Heather's bedroom. After just two
minutes, he came out of the bedroom and confronted
Heather.
[VF – D] She was shouting in a raspy voice: „I can‟t
believe how messy you are! The first thing I found on
the floor was your dirty pyjamas!‟
[VF – I] She was shouting in a raspy voice that she
could not believe how messy he was, and that the first
thing she found on the floor was his dirty pyjamas.
[VU – D] She said, pointing at the bedroom door: „I
can‟t believe how messy you are! The first thing I
found on the floor was your dirty pyjamas!‟
[VU – I] She said, pointing at the bedroom door, that
she could not believe how messy he was, and that the
first thing she found on the floor was his dirty
pyjamas.
Morton needed to speak to his project supervisor
Marianne at her office. When he asked whether she
could spare a minute, she seemed rather busy.
[VF – D] Ben was shouting in a raspy voice: „I can‟t
believe how messy you are! The first thing I found on
the floor was your dirty pyjamas!‟
[VF – I] Ben was shouting in a raspy voice that he
could not believe how messy she was, and that the first
thing he found on the floor was her dirty pyjamas.
[VU – D] He said, pointing at the bedroom door: „I
can‟t believe how messy you are! The first thing I
found on the floor was your dirty pyjamas!‟
[VU – I] He said, pointing at the bedroom door, that
he could not believe how messy she was, and that the
first thing he found on the floor was her dirty pyjamas.
[VF – D] She replied in a gentle and apologetic tone:
„I‟m really sorry, but I have to first finish marking
some essays.‟
[VF – I] She replied in a gentle and apologetic tone
that she was really sorry and that she had to first finish
marking some essays.
[VU – D] She replied, with an apologetic look on her
face: „I‟m really sorry, but I have to first finish
marking some essays.‟
[VU – I] She replied, with an apologetic look on her
face, that she was really sorry and that she had to first
finish marking some essays.
[VF – D] He replied in a gentle and apologetic tone:
„I‟m really sorry, but I have to first finish marking
some essays.‟
[VF – I] He replied in a gentle and apologetic tone
that he was really sorry and that he had to first finish
marking some essays.
[VU – D] He replied, with an apologetic look on his
face: „I‟m really sorry, but I have to first finish
marking some essays.‟
[VU – I] He replied, with an apologetic look on his
face, that he was really sorry and that he had to first
finish marking some essays.
Marianne needed to speak to her project supervisor
Morton at his office. When she asked whether he
could spare a minute, he seemed rather busy.
Page | 160
28
29
30
It was the Christmas Eve and everyone was drinking
and dancing at the local pub. Judith was already quite
tipsy when she noticed that George was with a girl she
had not seen before.
It was the Christmas Eve and everyone was drinking
and dancing at the local pub. George was already quite
tipsy when he noticed that Judith was with a girl he
had not seen before.
[VF – D] She went over, and in a playful voice she
said: „George, you‟ve actually never told me that you
had such a beautiful sister.‟
[VF – I] She went over, and in a playful voice she said
that he actually never told her that he had such a
beautiful sister.
[VU – D] She went over, smiled, and said: „George,
you‟ve actually never told me that you had such a
beautiful sister.‟
[VU – I] She went over, smiled, and said that he
actually never told her that he had such a beautiful
sister.
Maureen had helped David a lot with his coursework,
so David wanted to invite her for dinner. Asked what
kind of food she would fancy, Maureen first thought
for a second.
[VF – D] He went over, and in a playful voice he said:
„Jude, you‟ve actually never told me that you had such
a beautiful sister.‟
[VF – I] He went over, and in a playful voice he said
that she actually never told him that she had such a
beautiful sister.
[VU – D] He went over, smiled, and said: „Jude,
you‟ve actually never told me that you had such a
beautiful sister.‟
[VU – I] He went over, smiled, and said that she
actually never told him that she had such a beautiful
sister.
David had helped Maureen a lot with her coursework,
so Maureen wanted to invite him for dinner. Asked
what kind of food he would fancy, David first thought
for a second.
[VF – D] Then she replied, with some hesitation in her
voice: „I‟m normally not that adventurous, but perhaps
I should try some sushi.‟
[VF – I] Then she replied, with some hesitation in her
voice, that although she would normally not be that
adventurous, perhaps she should try some sushi.
[VU – D] Then she replied, with an indecisive look on
her face: „I‟m normally not that adventurous, but
perhaps I should try some sushi.‟
[VU – I] Then she replied, with an indecisive look on
her face, that although she would normally not be that
adventurous, perhaps she should try some sushi.
Jane was a new post-graduate at the department. When
she signed up at the Sports Club, she was having a
conversation with Gregg about her favourite sports.
[VF – D] Then he replied, with some hesitation in his
voice: „I‟m normally not that adventurous, but perhaps
I should try some sushi.‟
[VF – I] Then he replied, with some hesitation in his
voice, that although he would normally not be that
adventurous, perhaps he should try some sushi.
[VU – D] Then he replied, with an indecisive look on
his face: „I‟m normally not that adventurous, but
perhaps I should try some sushi.‟
[VU – I] Then he replied, with an indecisive look on
his face, that although he would normally not be that
adventurous, perhaps he should try some sushi.
Gregg was a new post-graduate at the department.
When he signed up at the Sports Club, he was having a
conversation with Jane about his favourite sports.
[VF – D] She said, not without pride in her voice: „Oh,
I haven‟t practised a lot recently, but I used to be quite
good at tennis.‟
[VF – I] She said, not without pride in her voice, that
although she had not practised a lot recently, she used
to be quite good at tennis.
[VU – D] She said, mocking a tennis serve: „Oh, I
haven‟t practised a lot recently, but I used to be quite
good at tennis.‟
[VU – I] She said, mocking a tennis serve, that
although she had not practised a lot recently, she used
to be quite good at tennis.
[VF – D] He said, not without pride in his voice: „Oh,
I haven‟t practised a lot recently, but I used to be quite
good at tennis.‟
[VF – I] He said, not without pride in his voice, that
although he had not practised a lot recently, he used to
be quite good at tennis.
[VU – D] He said, mocking a tennis serve: „Oh, I
haven‟t practised a lot recently, but I used to be quite
good at tennis.‟
[VU – I] He said, mocking a tennis serve, that
although he had not practised a lot recently, he used to
be quite good at tennis.
Page | 161
31
32
33
Betty and Leon were trying to get a mortgage for a
new flat. While Leon was studying the relevant
newspaper adverts, Betty was browsing some websites.
Leon and Betty were trying to get a mortgage for a
new flat. While Betty was studying the relevant
newspaper adverts, Leon was browsing some websites.
[VF – D] After a while, she sounded quite
disillusioned: „Hmm, these days, it looks as if nobody
would give us an instant cashback.‟
[VF – I] After a while, she sounded quite
disillusioned, saying that it looked as if nobody would
give them an instant cashback.
[VU – D] After a while, she sipped her tea and said:
„Hmm, these days, it looks as if nobody would give us
an instant cashback.‟
[VU – I] After a while, she sipped her tea and said that
it looked as if nobody would give them an instant
cashback.
Helen and Marcus were planning a trip to Europe for
their wedding anniversary. Helen was leaning more
towards Prague although Marcus preferred Rome.
[VF – D] After a while, he sounded quite
disillusioned: „Hmm, these days, it looks as if nobody
would give us an instant cashback.‟
[VF – I] After a while, he sounded quite disillusioned,
saying that it looked as if nobody would give them an
instant cashback.
[VU – D] After a while, he sipped his tea and said:
„Hmm, these days, it looks as if nobody would give us
an instant cashback.‟
[VU – I] After a while, he sipped his tea and said that
it looked as if nobody would give them an instant
cashback.
Marcus and Helen were planning a trip to Europe for
their wedding anniversary. Marcus was leaning more
towards Prague although Helen preferred Rome.
[VF – D] Helen said in a cheeky voice: „Okay, Rome
is nice, but the Czechs actually have much better
beer.‟
[VF – D] Marcus said in a cheeky voice: „Okay, Rome
is nice, but the Czechs actually have much better
beer.‟
[VF – I] Marcus said in a cheeky voice that Rome
would be nice, but that the Czechs would actually have
much better beer.
[VU – D] Marcus explained, with a cheeky smile:
„Okay, Rome is nice, but the Czechs actually have
much better beer.‟
[VU – I] Marcus explained, with a cheeky smile, that
Rome would be nice, but that the Czechs would
actually have much better beer.
[VF – I] Helen said in a cheeky voice that Rome
would be nice, but that the Czechs would actually have
much better beer.
[VU – D] Helen explained, with a cheeky smile:
„Okay, Rome is nice, but the Czechs actually have
much better beer.‟
[VU – I] Helen explained, with a cheeky smile, that
Rome would be nice, but that the Czechs would
actually have much better beer.
Dianne and Bill went out bowling together. When Bill
wanted to pick a ball of his favourite colour, Dianne
looked a bit worried.
[VF – D] She said, in a caring and considerate tone:
„That one looks quite filthy. I‟d rather pick the green
one which I think is much cleaner.‟
[VF – I] She said, in a caring and considerate tone,
that the ball looked quite filthy and that she‟d rather
pick the green one which she thought was much
cleaner.
[VU – D] She said, with a frown on her face: „That
one looks quite filthy. I‟d rather pick the green one
which I think is much cleaner.‟
[VU – I] She said, with a frown on her face, that the
ball looked quite filthy and that she‟d rather pick the
green one which she thought was much cleaner.
Bill and Dianne went out bowling together. When
Dianne wanted to pick a ball of her favourite colour,
Bill looked a bit worried.
[VF – D] He said, in a caring and considerate tone:
„That one looks quite heavy. I‟d rather pick the green
one which I think is much smaller.‟
[VF – I] He said, in a caring and considerate tone, that
the ball looked quite heavy and that he‟d rather pick
the green one which he thought was much smaller.
[VU – D] He said, with a frown on his face: „That one
looks quite heavy. I‟d rather pick the green one which
I think is much smaller.‟
[VU – I] He said, with a frown on his face, that the
ball looked quite heavy and that he‟d rather pick the
green one which he thought was much smaller.
Page | 162
34
35
36
Becky and Rob were buying duvet covers at the store.
When Rob chose a bright red one, Becky was
completely against the colour.
Rob and Becky were buying duvet covers at the store.
When Becky chose a bright red one, Rob was
completely against the colour.
[VF – D] In a piercing, sarcastic tone, she said: „No
way! Just imagine that every morning you'd be waking
up in a sea of blood!‟
[VF – I] In a piercing, sarcastic tone, she asked Rob
just to imagine that every morning he would be
waking up in a sea of blood.
[VU – D] She said, with her head shaking frantically:
„No way! Just imagine that every morning you'd be
waking up in a sea of blood!‟
[VU – I] She said, with her head shaking frantically,
just to imagine that every morning he would be
waking up in a sea of blood.
Leona and Harry were enjoying an afternoon walk in
the park. When they were approaching the pond,
Leona was impressed by this peaceful place.
[VF – D] In a piercing, sarcastic tone, he said: „No
way! Just imagine that every morning you'd be waking
up in a sea of blood!‟
[VF – I] In a piercing, sarcastic tone, he asked Becky
just to imagine that every morning she would be
waking up in a sea of blood.
[VU – D] He said, with his head shaking frantically:
„No way! Just imagine that every morning you'd be
waking up in a sea of blood!‟
[VU – I] He said, with his head shaking frantically,
just to imagine that every morning she would be
waking up in a sea of blood.
Harry and Leona were enjoying an afternoon walk in
the park. When they were approaching the pond, Harry
was impressed by this peaceful place.
[VF – D] Her voice sounded totally smitten and
cheerful: „This is so relaxing after a busy week! It is
almost magical.‟
[VF – I] Her voice sounded totally smitten and
cheerful, saying that it was so relaxing after a busy
week, and that it was almost magical.
[VU – D] She closed her eyes and took a deep breath:
„This is so relaxing after a busy week! It is almost
magical.‟
[VU – I] She closed her eyes and took a deep breath,
saying that it was so relaxing after a busy week, and
that it was almost magical.
Schoolteacher Lynda recently caused a stir by
marrying Edward, one of her ex-pupils. Annoyed by
all the gossip, Lynda told Edward to simply not listen
to what people say.
[VF – D] His voice sounded totally smitten and
cheerful: „This is so relaxing after a busy week! It is
almost magical.‟
[VF – I] His voice sounded totally smitten and
cheerful, saying that it was so relaxing after a busy
week, and that it was almost magical.
[VU – D] He closed his eyes and took a deep breath:
„This is so relaxing after a busy week! It is almost
magical.‟
[VU – I] He closed his eyes and took a deep breath,
saying that it was so relaxing after a busy week, and
that it was almost magical.
Schoolteacher Edward recently caused a stir by
marrying Lynda, one of his ex-pupils. Annoyed by all
the gossip, Edward told Lynda to simply not listen to
what people say.
[VF – D] She said, in a firm voice: „My dear, I think
there‟s nothing wrong about a teacher marrying a
former student.‟
[VF – I] She said, in a firm voice, that she thought
there was nothing wrong about a teacher marrying a
former student.
[VU – D] With a resolute expression on her face, she
said: „My dear, I think there‟s nothing wrong about a
teacher marrying a former student.‟
[VU – I] With a resolute expression on her face, she
said that she thought there was nothing wrong about a
teacher marrying a former student.
[VF – D] He said, in a firm voice: „My dear, I think
there‟s nothing wrong about a teacher marrying a
former student.‟
[VF – I] He said, in a firm voice, that he thought there
was nothing wrong about a teacher marrying a former
student.
[VU – D] With a resolute expression on his face, he
said: „My dear, I think there‟s nothing wrong about a
teacher marrying a former student.‟
[VU – I] With a resolute expression on his face, he
said that he thought there was nothing wrong about a
teacher marrying a former student.
Page | 163
37
38
39
After making a small fortune in the lottery, Ralph
bought quite an expensive car. Today in the car park,
his colleague Pamela was extremely impressed by his
new possession.
After making a small fortune in the lottery, Pamela
bought quite an expensive car. Today in the car park,
her colleague Ralph was extremely impressed by her
new possession.
[VF – D] Her voice wavered between admiration and
envy: „Wow! Is that your new car? It looks really
stylish.‟
[VF – I] Her voice wavered between admiration and
envy when she asked him whether that was his new
car, adding that it looked really stylish.
[VU – D] Her face wavered between admiration and
envy: „Wow! Is that your new car? It looks really
stylish.‟
[VU – I] Her face wavered between admiration and
envy when she asked him whether that was his new
car, adding that it looked really stylish.
Katy invited some of her friends home and started
making a mess in the kitchen. Her mum was obviously
not happy about this, and Katy was quick to offer an
excuse.
[VF – D] His voice wavered between admiration and
envy: „Wow! Is that your new car? It looks really
stylish.‟
[VF – I] His voice wavered between admiration and
envy when he asked her whether that was her new car,
adding that it looked really stylish.
[VU – D] His face wavered between admiration and
envy: „Wow! Is that your new car? It looks really
stylish.‟
[VU – I] His face wavered between admiration and
envy when he asked her whether that was her new car,
adding that it looked really stylish.
Brian invited some of his of mates home and started
making a mess in the kitchen. His mum was obviously
not happy about this, and Brian was quick to offer an
excuse.
[VF – D] Her voice sounded apprehensive: „Well, we
just had a couple of drinks because the pubs are
already closed.‟
[VF – I] Her voice sounded apprehensive, saying that
they just had a couple of drinks because the pubs were
already closed.
[VU – D] She was staring at the floor: „Well, we just
had a couple of drinks because the pubs are already
closed.‟
[VU – I] She was staring at the floor, saying that they
just had a couple of drinks because the pubs were
already closed.
Maddie was injured in a car crash last week. At the
hospital, she was quickly recovering. Today, she was
surprised that some distant relatives came to visit her.
[VF – D] His voice sounded apprehensive: „Well, we
just had a couple of drinks because the pubs are
already closed.‟
[VF – I] His voice sounded apprehensive, saying that
they just had a couple of drinks because the pubs were
already closed.
[VU – D] He was staring at the floor: „Well, we just
had a couple of drinks because the pubs are already
closed.‟
[VU – I] He was staring at the floor, saying that they
just had a couple of drinks because the pubs were
already closed.
Daniel was injured in a car crash last week. At the
hospital, he was quickly recovering. Today, he was
surprised that some distant relatives came to visit him.
[VF – D] Maddie sounded sarcastic: „Oh, looks as if
one has to get hurt to get some flowers.‟
[VF – I] Maddie sounded sarcastic, saying that it
looked as if one had to get hurt to get some flowers.
[VU – D] Maddie said, with a sarcastic smile on her
face: „Oh, looks as if one has to get hurt to get some
flowers.‟
[VU – I] Maddie said, with a sarcastic smile on her
face, that it looked as if one had to get hurt to get some
flowers.
[VF – D] Daniel sounded sarcastic: „Oh, looks as if
one has to get hurt to get some flowers.‟
[VF – I] Daniel sounded sarcastic, saying that it
looked as if one had to get hurt to get some flowers.
[VU – D] Daniel said, with a sarcastic smile on his
face: „Oh, looks as if one has to get hurt to get some
flowers.‟
[VU – I] Daniel said, with a sarcastic smile on his
face, that it looked as if one had to get hurt to get some
flowers.
Page | 164
40
41
42
43
Cindy had promised Jacob to buy a bonsai tree to
decorate their living room. When Jacob wanted to
place the plant by the television, she tried to persuade
him of an alternative option.
Jacob had promised Cindy to buy a bonsai tree to
decorate their living room. When Cindy wanted to
place the plant by the television, he tried to persuade
her of an alternative option.
[VF – D] She explained in a clear and firm voice:
„Look, if you do that, there will be no sunlight for the
bonsai, so it's better to place it by the window.‟
[VF – I] She explained in a clear and firm voice that if
he did that, there would be no sunlight for the bonsai,
so it was better to place it by the window.
[VU – D] She explained, with a wagging forefinger:
„Look, if you do that, there will be no sunlight for the
bonsai, so it's better to place it by the window.‟
[VU – I] She explained, with a wagging forefinger,
that if he did that, there would be no sunlight for the
bonsai, so it was better to place it by the window.
PhD student Jim was summoned to her supervisor
Ella‟s office to give a report on his current progress.
When Jim asked for an extension, Ella looked
concerned.
[VF – D] He explained in a clear and firm voice:
„Look, if you do that, there will be no sunlight for the
bonsai, so it's better to place it by the window.‟
[VF – I] He explained in a clear and firm voice that if
she did that, there would be no sunlight for the bonsai,
so it was better to place it by the window.
[VU – D] He explained, with a wagging forefinger:
„Look, if you do that, there will be no sunlight for the
bonsai, so it's better to place it by the window.‟
[VU – I] He explained, with a wagging forefinger, that
if she did that, there would be no sunlight for the
bonsai, so it was better to place it by the window.
PhD student Ella was summoned to her supervisor
Jim‟s office to give a report on her current progress.
When Ella asked for an extension, Jim looked
concerned.
[VF – D] Her voice turned serious: „Hmm, we really
need those data in by next month for that conference.‟
[VF – I] Her voice turned serious, saying that they
really needed those data in by next month for that
conference.
[VU – D] Her face turned serious: „Hmm, we really
need those data in by next month for that conference.‟
[VU – I] Her face turned serious, saying that they
really needed those data in by next month for that
conference.
Joanne was taking Albert to the new Schwarzenegger
movie. Joanne was very excited, but since Albert
seemed not very keen, she tried to persuade him to go
with her.
[VF – D] His voice turned serious: „Hmm, we really
need those data in by next month for that conference.‟
[VF – I] His voice turned serious, saying that they
really needed those data in by next month for that
conference.
[VU – D] His face turned serious: „Hmm, we really
need those data in by next month for that conference.‟
[VU – I] His face turned serious, saying that they
really needed those data in by next month for that
conference.
Albert was taking Joanne to the new Schwarzenegger
movie. Albert was very excited, but since Joanne
seemed not very keen, he tried to persuade her to go
with him.
[VF – D] She raised her voice, sounding very agitated:
„Come on! I‟m sure you will love it! The movie has
some really cool special effects.‟
[VF – I] She raised her voice, sounding very agitated,
and said that she was sure he would love it, and that
the movie had some really cool special effects.
[VU – D] She said, with her eyes wide open: „Come
on! I‟m sure you will love it! The movie has some
really cool special effects.‟
[VU – I] She said, with her eyes wide open, that she
was sure he would love it, and that the movie had
some really cool special effects.
Stephanie and Neil were visiting the local distillery.
Stephanie was amazed by the range of different
Whiskeys produced, and she tried a very expensive
Single Malt.
[VF – D] She coughed and said in a breathy voice:
„Gee, that‟s a strong one! I‟m glad I won‟t have to
drive us home.‟
[VF – I] She coughed and said in a breathy voice that
it was really strong and that she was glad she didn‟t
have to drive them home.
[VU – D] She said, slapping her chest: „Gee, that‟s a
strong one! I‟m glad I won‟t have to drive us home.‟
[VU – I] She said, slapping her chest, that it was really
strong and that she was glad she didn‟t have to drive
them home.
[VF – D] He raised his voice, sounding very agitated:
„Come on! I‟m sure you will love it! The movie has
some really cool special effects.‟
[VF – I] He raised his voice, sounding very agitated,
and said that he was sure she would love it, and that
the movie had some really cool special effects.
[VU – D] He said, with his eyes wide open: „Come on!
I‟m sure you will love it! The movie has some really
cool special effects.‟
[VU – I] He said, with his eyes wide open, that he was
sure she would love it, and that the movie had some
really cool special effects.
Neil and Stephanie were visiting the local distillery.
Neil was amazed by the range of different Whiskeys
produced, and he tried a very expensive Single Malt.
[VF – D] He coughed and said in a breathy voice:
„Gee, that‟s a strong one! I‟m glad I won‟t have to
drive us home.‟
[VF – I] He coughed and said in a breathy voice that it
was really strong and that he was glad he didn‟t have
to drive them home.
[VU – D] He said, slapping his chest: „Gee, that‟s a
strong one! I‟m glad I won‟t have to drive us home.‟
[VU – I] He said, slapping his chest, that it was really
strong and that he was glad he didn‟t have to drive
them home.
Page | 165
44
45
46
47
Molly and her husband Clive went out to the new
seafood restaurant for the first time. Molly was
extremely anxious about food poisoning and did not
want to have anything raw.
Clive and his wife Molly went out to the new seafood
restaurant for the first time. Clive was extremely
anxious about food poisoning and did not want to have
anything raw.
[VF – D] In a firm voice, she said to Clive: „I don‟t
care what other people might think, but I will order
properly cooked oysters.‟
[VF – I] In a firm voice, she said to Clive that,
regardless of what other people might think, she would
order properly cooked oysters.
[VU – D] With a serious look on her face, she said to
Clive: „I don‟t care what other people might think, but
I will order properly cooked oysters.‟
[VU – I] With a serious look on her face, she said to
Clive that regardless of what other people might think,
she would order properly cooked oysters.
Lucy and her friends were watching a movie at the
cinema. Lucy wasn‟t particularly keen on romantic
comedies, and she was complaining a lot after the
film.
[VF – D] In a firm voice, he said to Molly: „I don‟t
care what other people might think, but I will order
properly cooked oysters.‟
[VF – I] In a firm voice, he said to Molly that,
regardless of what other people might think, he would
order properly cooked oysters.
[VU – D] With a serious look on his face, he said to
Molly: „I don‟t care what other people might think, but
I will order properly cooked oysters.‟
[VU – I] With a serious look on his face, he said to
Molly that regardless of what other people might
think, he would order properly cooked oysters.
Luke and his friends were watching a movie at the
cinema. Luke wasn‟t particularly keen on romantic
comedies, and he was complaining a lot after the film.
[VF – D] Her voice sounded harsh and uncontent:
„God, that movie was terrible! I‟ve never been so
bored in my life.‟
[VF – I] Her voice sounded harsh and uncontent,
saying that the movie was terrible and that she had
never been so bored in her life.
[VU – D] She was rolling her eyes: „God, that movie
was terrible! I‟ve never been so bored in my life.‟
[VU – I] She was rolling her eyes, saying that the
movie was terrible and that she had never been so
bored in her life.
On the train to London, Thomas was very nervous
about his upcoming job interview. When he reached
for a cigarette to calm himself down, the girl next to
him complained instantly.
[VF – D] His voice sounded harsh and uncontent:
„God, that movie was terrible! I‟ve never been so
bored in my life.‟
[VF – I] His voice sounded harsh and uncontent,
saying that the movie was terrible and that he had
never been so bored in his life.
[VU – D] He was rolling his eyes: „God, that movie
was terrible! I‟ve never been so bored in my life.‟
[VU – I] He was rolling his eyes, saying that the
movie was terrible and that he had never been so bored
in his life.
On the train to London, Thomas was very nervous
about his upcoming job interview. When he reached
for a cigarette to calm himself down, the middle-aged
man next to him complained instantly.
[VF – D] The girl called out, in a stroppy tone: „Oi!
Smoking on public transport is illegal!‟
[VF – I] The girl called out, in a stroppy tone, that
smoking on public transport was illegal.
[VU – D] The girl called out, with an angry look on
her face: „Oi! Smoking on public transport is illegal!‟
[VU – I] The girl called out, with an angry look on her
face, that smoking on public transport was illegal.
During a coffee break, university teachers Isabel and
Robert met in the common room. Isabel was slightly
irritated with the mess in the kitchen.
[VF – D] The man called out, in a stroppy tone: „Oi!
Smoking on public transport is illegal!‟
[VF – I] The man called out, in a stroppy tone, that
smoking on public transport was illegal.
[VU – D] The man called out, with an angry look on
his face: „Oi! Smoking on public transport is illegal!‟
[VU – I] The man called out, with an angry look on
his face, that smoking on public transport was illegal.
During a coffee break, university teachers Robert and
Isabel met in the common room. Robert was slightly
irritated with the mess in the kitchen.
[VF – D] Her voice was indignant: „Look at this!
Certain people must always leave their dirty mugs on
the table.‟
[VF – I] Her voice was indignant, saying that,
obviously, certain people must always leave their dirty
mugs on the table.
[VU – D] She said, pointing at the kitchen table:
„Look at this! Certain people must always leave their
dirty mugs on the table.‟
[VU – I] She said, pointing at the kitchen table, that
obviously, certain people must always leave their dirty
mugs on the table.
[VF – D] His voice was indignant: „Look at this!
Certain people must always leave their dirty mugs on
the table.‟
[VF – I] His voice was indignant, saying that,
obviously, certain people must always leave their dirty
mugs on the table.
[VU – D] He said, pointing at the kitchen table: „Look
at this! Certain people must always leave their dirty
mugs on the table.‟
[VU – I] He said, pointing at the kitchen table, that
obviously, certain people must always leave their dirty
mugs on the table.
Page | 166
48
49
50
School teachers Liana and Duncan were arguing about
the curriculum. Apparently, Liana was not a big fan of
Duncan‟s modern style of teaching.
School teachers Duncan and Liana were arguing about
the curriculum. Apparently, Duncan was not a big fan
of Liana‟s modern style of teaching.
[VF – D] She sounded arrogant and irreverent: „If my
own children were at this school, I‟d make sure they
would never have to attend your lessons.‟
[VF – I] She sounded arrogant and irreverent, saying
that if her own children were at this school, she‟d
make sure they would never have to attend his lessons.
[VU – D] She said, with an arrogant look on her face:
„If my own children were at this school, I‟d make sure
they would never have to attend your lessons.‟
[VU – I] She said, with an arrogant look on her face,
that if her own children were at this school, she‟d
make sure they would never have to attend his lessons.
[VF – D] He sounded arrogant and irreverent: „If my
own children were at this school, I‟d make sure they
would never have to attend your lessons.‟
[VF – I] He sounded arrogant and irreverent, saying
that if his own children were at this school, he‟d make
sure they would never have to attend her lessons.
It was getting quite late and most of Scott‟s friends
were about to leave his party. Andrea in particular
seemed to have had a great time when she was
leaving.
[VF – D] Her voice was jolly and cheerful: „Sorry
about the crack in that glass. The party was absolutely
fantastic.‟
[VF – I] Her voice was jolly and cheerful when she
apologised for the crack in that glass, adding that the
party was absolutely fantastic.
[VU – D] She apologised, with a big smile: „Sorry
about the crack in that glass. The party was absolutely
fantastic.‟
[VU – I] She apologised, with a big smile, for the
crack in that glass, adding that the party was
absolutely fantastic.
Donna always enjoyed a bit of gardening in her spare
time. She recently bought a new lawnmower, but she
was quite frustrated because it simply refused to work.
[VF – D] She complained loudly and resentfully: „Sod
it! Things used to be much better in the past. All one
can buy these days is rubbish!‟
[VF – I] She complained loudly and resentfully that
things used to be much better in the past, and that all
one can buy these days was rubbish.
[VU – D] She complained with a very long face: „Sod
it! Things used to be much better in the past. All one
can buy these days is rubbish!‟
[VU – I] She complained with a very long face that
things used to be much better in the past, and that all
one can buy these days was rubbish.
[VU – D] He said, with an arrogant look on his face:
„If my own children were at this school, I‟d make sure
they would never have to attend your lessons.‟
[VU – I] He said, with an arrogant look on his face,
that if his own children were at this school, he‟d make
sure they would never have to attend her lessons.
It was getting quite late and most of Andrea‟s friends
were about to leave her party. Scott in particular
seemed to have had a great time when he was leaving.
[VF – D] His voice was jolly and cheerful: „Sorry
about the crack in that glass. The party was absolutely
fantastic.‟
[VF – I] His voice was jolly and cheerful when he
apologised for the crack in that glass, adding that the
party was absolutely fantastic.
[VU – D] He apologised, with a big smile: „Sorry
about the crack in that glass. The party was absolutely
fantastic.‟
[VU – I] He apologised, with a big smile, for the crack
in that glass, adding that the party was absolutely
fantastic.
Doug always enjoyed a bit of gardening in his spare
time. He recently bought a new lawnmower, but he
was quite frustrated because it simply refused to work.
[VF – D] He complained loudly and resentfully: „Sod
it! Things used to be much better in the past. All one
can buy these days is rubbish!‟
[VF – I] He complained loudly and resentfully that
things used to be much better in the past, and that all
one can buy these days was rubbish.
[VU – D] He complained with a very long face: „Sod
it! Things used to be much better in the past. All one
can buy these days is rubbish!‟
[VU – I] He complained with a very long face that
things used to be much better in the past, and that all
one can buy these days was rubbish.
Page | 167
51
52
53
After class, Rachel had a serious conversation with
one of his pupils. The boy had been accused of
bullying and Rachel was giving him a caution.
After class, Roy had a serious conversation with one
of his pupils. The boy had been accused of bullying
and Roy was giving him a caution.
[VF – D] Rachel sounded very formal and serious: „If
the accusations are true, you will be in serious trouble.
One cannot tolerate such kind of behaviour.‟
[VF – I] Rachel sounded very formal and serious,
saying that if the accusations were true, the boy would
be in serious trouble, and that one cannot tolerate such
kind of behaviour.
[VU – D] Rachel said, with a serious expression on
her face: „If the accusations are true, you will be in
serious trouble. One cannot tolerate such kind of
behaviour.‟
[VU – I] Rachel said, with a serious expression on her
face, that if the accusations were true, the boy would
be in serious trouble, and that one cannot tolerate such
kind of behaviour.
In the board room, assistant manager Chloe was giving
a presentation on how to improve the company‟s sales
figures. When she noticed that her boss was anything
but impressed, Chloe used her lack of experience as an
excuse.
[VF – D] Roy sounded very formal and serious: „If the
accusations are true, you will be in serious trouble.
One cannot tolerate such kind of behaviour.‟
[VF – I] Roy sounded very formal and serious, saying
that if the accusations were true, the boy would be in
serious trouble, and that one cannot tolerate such kind
of behaviour.
[VU – D] Roy said, with a serious expression on his
face: „If the accusations are true, you will be in serious
trouble. One cannot tolerate such kind of behaviour.‟
[VU – I] Roy said, with a serious expression on his
face, that if the accusations were true, the boy would
be in serious trouble, and that one cannot tolerate such
kind of behaviour.
[VF – D] Her voice was shy and hesitant: „Sorry, this
is actually my first ever presentation and I‟m probably
not very persuasive.‟
[VF – I] Her voice was shy and hesitant, saying that
this was actually her first ever presentation, and that
she was probably not very persuasive.
[VU – D] She said, looking rather nervous: „Sorry, this
is actually my first ever presentation and I‟m probably
not very persuasive.‟
[VU – I] She said, looking rather nervous, that this
was actually her first ever presentation, and that she
was probably not very persuasive.
Smith was working in a small antiques shop down the
local high street. Today, a middle-aged posh lady with
thick glasses came into the shop.
[VF – D] His voice was shy and hesitant: „Sorry, this
is actually my first ever presentation and I‟m probably
not very persuasive.‟
[VF – I] His voice was shy and hesitant, saying that
this was actually his first ever presentation, and that he
was probably not very persuasive.
[VU – D] He said, looking rather nervous: „Sorry, this
is actually my first ever presentation and I‟m probably
not very persuasive.‟
[VU – I] He said, looking rather nervous, that this was
actually his first ever presentation, and that he was
probably not very persuasive.
Alice was working in a small antiques shop down the
local high street. Today, a weird-looking man with
greasy hair and thick glasses came into the shop.
[VF – D] She looked around and said in a nonchalant
tone: „You may be surprised to learn that I‟m a worldrenowned collector of rare memorabilia of Whiteeared Pheasant.‟
[VF – I] She looked around and said, in a nonchalant
tone, that he might be surprised to learn that she was a
world-renowned collector of rare memorabilia of
White-eared Pheasant.
[VU – D] She looked around and said with a slightly
absent facial expression: „You may be surprised to
learn that I‟m a world-renowned collector of rare
memorabilia of White-eared Pheasant.‟
[VU – I] She looked around and said, with a slightly
absent facial expression, that he might be surprised to
learn that she was a world-renowned collector of rare
memorabilia of White-eared Pheasant.
[VF – D] He looked around and said in a slightly
croaky voice: „You may be surprised to learn that I‟m
a world-renowned collector of rare memorabilia of
White-eared Pheasant.‟
[VF – I] He looked around and said, in a slightly
croaky voice, that she might be surprised to learn that
he was a world-renowned collector of rare
memorabilia of White-eared Pheasant.
[VU – D] He looked around and said with a slightly
absent facial expression: „You may be surprised to
learn that I‟m a world-renowned collector of rare
memorabilia of White-eared Pheasant.‟
[VU – I] He looked around and said, with a slightly
absent facial expression, that she might be surprised to
learn that he was a world-renowned collector of rare
memorabilia of White-eared Pheasant.
In the board room, assistant manager Craig was giving
a presentation on how to improve the company‟s sales
figures. When he noticed that his boss was anything
but impressed, Craig used his lack of experience as an
excuse.
Page | 168
54
55
56
57
Sandra and Ray were on a trip to the countryside for
the weekend. Sandra quite liked the outdoors and
particularly that rural smell.
Ray and Sandra were on a trip to the countryside for
the weekend. Ray quite liked the outdoors and
particularly that rural smell.
[VF – D] She said in a relaxed and content voice:
„Aah, this is so much better than crowded cities and
pollution.‟
[VF – I] She said, in a relaxed and content voice, that
it was so much better than crowded cities and
pollution.
[VU – D] She said, with a content smile: „Aah, this is
so much better than crowded cities and pollution.‟
[VU – I] She said, with a content smile, that it was so
much better than crowded cities and pollution.
Earlier in the afternoon, Janis and Max were having a
cigarette in front of the main entrance. Not knowing
what to say, Janis started her typical conversation
about the weather.
[VF – D] He said in a relaxed and content voice: „Aah,
this is so much better than crowded cities and
pollution.‟
[VF – I] He said, in a relaxed and content voice, that it
was so much better than crowded cities and pollution.
[VU – D] He said, with a content smile: „Aah, this is
so much better than crowded cities and pollution.‟
[VU – I] He said, with a content smile, that it was so
much better than crowded cities and pollution.
[VF – D] She tried to speak in a witty tone: „Gosh, it‟s
quite windy today – but at least it‟s not raining.‟
[VF – I] She tried to speak in a witty tone, saying that
although it was quite windy today, at least it was not
raining.
[VU – D] She said, puffing away her cigarette: „Gosh,
it‟s quite windy today – but at least it‟s not raining.‟
[VU –I] She said, puffing away her cigarette, that
although it was quite windy today, at least it was not
raining.
A journalist was interviewing Amy, the older sister of
a famous actor. Amy did not really enjoy the attention
she was given and wasn‟t sure what to say.
[VF- D] He tried to speak in a witty tone: „Gosh, it‟s
quite windy today – but at least it‟s not raining.‟
[VF – I] He tried to speak in a witty tone, saying that
although it was quite windy today, at least it was not
raining.
[VU – D] He said, puffing away his cigarette: „Gosh,
it‟s quite windy today – but at least it‟s not raining.‟
[VU – I] He said, puffing away his cigarette, that
although it was quite windy today, at least it was not
raining.
A journalist was interviewing Eric, the older brother of
a famous pop diva. Eric did not really enjoy the
attention he was given and wasn‟t sure what to say.
[VF – D] She proclaimed in a slightly annoyed tone: „I
actually don‟t see my brother very often these days.
He is completely devoted to his career!‟
[VF – I] She proclaimed, in a slightly annoyed tone,
that she actually did not see her brother very often in
those days, and that he was completely devoted to his
career.
[VU – D] She proclaimed, inspecting her fingernails:
„I actually don‟t see my brother very often these days.
He is completely devoted to his career!‟
[VU – I] She proclaimed, inspecting her fingernails,
that she actually did not see her brother very often in
those days, and that he was completely devoted to his
career.
Sheena went to the local McDonalds and incidentally
met Thomas, who he knew from former Weight
Watchers meetings. Naturally, they were both quite
embarrassed, but Sheena finally broke the silence.
[VF – D] He proclaimed in a slightly annoyed tone: „I
actually don‟t see my sister very often these days. She
is completely devoted to her career!‟
[VF – I] He proclaimed, in a slightly annoyed tone,
that he actually did not see his sister very often in
those days, and that she was completely devoted to her
career.
[VU – D] He proclaimed, inspecting his fingernails: „I
actually don‟t see my sister very often these days. She
is completely devoted to her career!‟
[VU – I] He proclaimed, inspecting his fingernails,
that he actually did not see his sister very often in
those days, and that she was completely devoted to her
career.
Thomas went to the local McDonalds and incidentally
met Sheena, who he knew from former Weight
Watchers meetings. Naturally, they were both quite
embarrassed, but Thomas finally broke the silence.
[VF – D] She tried to sound ironic: „Weight Watchers
is rubbish, isn‟t it? I‟ve tried so many diets but I‟m
still overweight.‟
[VF – I] She tried to sound ironic, saying that Weight
Watchers was rubbish, and that she had tried so many
diets but was still overweight.
[VU – D] She said, with an ironic wink: „Weight
Watchers is rubbish, isn‟t it? I‟ve tried so many diets
but I‟m still overweight.‟
[VU – I] She said, with an ironic wink, that Weight
Watchers was rubbish, and that she had tried so many
diets but was still overweight.
[VF – D] He tried to sound ironic: „Weight Watchers
is rubbish, isn‟t it? I‟ve tried so many diets but I‟m
still overweight.‟
[VF – I] He tried to sound ironic, saying that Weight
Watchers was rubbish, and that he had tried so many
diets but was still overweight.
[VU – D] He said, with an ironic wink: „Weight
Watchers is rubbish, isn‟t it? I‟ve tried so many diets
but I‟m still overweight.‟
[VU – I] He said, with an ironic wink, that Weight
Watchers was rubbish, and that he had tried so many
diets but was still overweight.
Earlier in the afternoon, Max and Janis were having a
cigarette in front of the main entrance. Not knowing
what to say, Max started his typical conversation about
the weather.
Page | 169
58
59
60
A car-boot sale was taking place in the neighbourhood
and Catherine was trying to get rid of some of the old
records she owned since the 1970s.A trendy young
man appeared to be very interested, so Catherine
started advertising.
A car-boot sale was taking place in the neighbourhood
and Kenny was trying to get rid of some of the old
records he owned since the 1970s.A trendy young
woman appeared to be very interested, so Kenny
started advertising.
[VF – D] She said in an enthusiastic and persuasive
tone: „I can give you three albums for a fiver. You
won‟t find any of that stuff on the internet.‟
[VF – I] She said in an enthusiastic and persuasive
tone that she could give him three albums for a fiver,
adding that he wouldn‟t find any of that stuff on the
internet.
[VU – D] She said, holding a stack of CDs in her
hands: „I can give you three albums for a fiver. You
won‟t find any of that stuff on the internet.‟
[VU – I] She said, holding a stack of CDs in her
hands, that she could give him three albums for a fiver,
adding that he wouldn‟t find any of that stuff on the
internet.
A trendy night club had been recently opened near
Patricia‟s flat and it proved instantly popular.
However, Patricia often felt disturbed by the noise,
and this morning she was complaining to the
neighbours about it.
[VF – D] He said in an enthusiastic and persuasive
tone: „I can give you three albums for a fiver. You
won‟t find any of that stuff on the internet.‟
[VF – I] He said in an enthusiastic and persuasive tone
that he could give her three albums for a fiver, adding
that she wouldn‟t find any of that stuff on the internet.
[VU – D] He said, holding a stack of CDs in his
hands: „I can give you three albums for a fiver. You
won‟t find any of that stuff on the internet.‟
[VU – I] He said, holding a stack of CDs in his hands,
that he could give her three albums for a fiver, adding
that she wouldn‟t find any of that stuff on the internet.
[VF – D] Her voice was furious and raspy: „This is
unbelievable! Every time I have to work early hours,
they are having a bash!‟
[VF – I] Her voice was furious and raspy, saying that
every time she has to work early hours, they are
having a bash.
[VU – D] She said, with her face turning red: „This is
unbelievable! Every time I have to work early hours,
they are having a bash!‟
[VU – I] She said, with her face turning red, that every
time she has to work early hours, they are having a
bash.
Carol, a taxi driver, was about to drive home when
suddenly she saw a man by the road frantically waving
his arms. Carol stopped and the man offered her
double the fare if they make it to the airport within ten
minutes.
[VF – D] His voice was furious and raspy: „This is
unbelievable! Every time I have to work early hours,
they are having a bash!‟
[VF – I] His voice was furious and raspy, saying that
every time he has to work early hours, they are having
a bash.
[VU – D] He said, with his face turning red: „This is
unbelievable! Every time I have to work early hours,
they are having a bash!‟
[VU – I] He said, with his face turning red, that every
time he has to work early hours, they are having a
bash.
Carl, a taxi driver, was about to drive home when
suddenly he saw a man by the road frantically waving
his arms. Carl stopped and the man offered him double
the fare if they make it to the airport within ten
minutes.
[VF – D] Being professional, she replied in a calm and
firm voice: „OK. I‟ll try my best, mate, but cannot
work any miracles.‟
[VF – I] Being professional, she replied in a calm and
firm voice that she will try her best, but that she cannot
work any miracles.
[VU – D] Being professional, she replied with a calm
expression on her face: „OK. I‟ll try my best, mate, but
cannot work any miracles.‟
[VU – I] Being professional, she replied with a calm
expression on her face, saying that she will try her
best, but that she cannot work any miracles.
[VF – D] Being professional, Carl replied in a calm
and firm voice: „OK. I‟ll try my best, mate, but cannot
work any miracles.‟
[VF – I] Being professional, Carl replied in a calm and
firm voice that he will try his best, but that he cannot
work any miracles.
[VU – D] Being professional, Carl replied with a calm
expression on his face: „OK. I‟ll try my best, mate, but
cannot work any miracles.‟
[VU – I] Being professional, Carl replied with a calm
expression on his face, saying that he will try his best,
but that he cannot work any miracles.
A trendy night club had been recently opened near
Marvin‟s flat and it proved instantly popular.
However, Marvin often felt disturbed by the noise, and
this morning he was complaining to the neighbours
about it.
Page | 170
Appendix 3 Loudness stimuli
[D] = direct speech; [I] = indirect speech
The critical sentences are in bold and italic text (for acoustic analyses in oral reading).
Item
1
2
3
4
Loud-speaking story
Diane and Sharon entered the lobby during the
intermission and purchased a glass of wine. The room
was buzzing with hundreds of conversations about the
performance. Diane spotted her sister‟s cheating exhusband standing nearby with his new girlfriend and
intended to publicly embarrass him.
Quiet-speaking story
The orchestra were playing a delicate, melancholy
piece when Diane spotted her friend‟s cheating exhusband sitting nearby. She was fuming to see him and
nudged her sister, pointing him out to her subtly.
[D] She turned to Sharon and purposely increased her
voice: „Well! I can't believe he has the nerve to show
his face in public after his awful behaviour.‟
[I] She turned to Sharon and purposely increased her
voice saying that she couldn't believe he had the
nerve to show his face in public after his awful
behaviour.
Katy‟s friends were having such a great time blasting
cheesy Christmas music in her flat that she barely
heard the knock at the door at 1am. She was expecting
another friend and was surprised to find a policeman
looking crossly over her shoulder at the party going on
inside.
[D] She turned and said in an undertone to Sharon:
„Well! I can't believe he has the nerve to show his
face in public after his awful behaviour.‟
[I] She turned and said in an undertone to Sharon that
she couldn't believe he had the nerve to show his face
in public after his awful behaviour.
[D] He looked very serious, demanding: „Excuse me
Madam. I suggest you keep the noise down as there's
been a complaint from your neighbours.‟
[I] He looked very serious, demanding that she should
keep the noise down as there had been a complaint
from her neighbours.
The construction site was bustling with activity and
the shrill radio rang out in between the whirring of
electric tools and reversing lorries. The builders and
joiners had been working hard all morning so Bill, the
foreman, gestured wildly to his team scattered around
the site.
[D] He politely said: „Excuse me Madam. I suggest
you keep the noise down as there's been a complaint
from your neighbours.‟
[I] He politely said to her that she should keep the
noise down as there had been a complaint from her
neighbours.
All that could be heard was the noise of picks on dry
soil as the archaeologists painstakingly concentrated
on the task. Bill was disappointed not to have found
anything so far in spite of their conscientious work. He
sighed softly and suggested taking a break to his coworker.
[D] He yelled: „Okay, I think it’s time to have a cup
of coffee.‟
[I] He yelled that he thought it was time to have a cup
of coffee.
On one terrifying night during WWII, the bombs were
dropped over the city by the enemy and the troops
could hear the explosions as they took cover in ruined
buildings. Brian put his hands over his ears and
screamed uncontrollably as a blast nearby made the
walls of their hideout shake violently.
[D] He muttered: „Okay, I think it’s time to have a
cup of coffee.‟
[I] He muttered that he thought it was time to have a
cup of coffee.
It was a perfectly calm day and Sarah and Brian
drifted along the coast in their sea kayak, spotting
local birds and wildlife. As they rounded a cliff, they
came across a sea eagle nearby and Brian turned
around and put his finger to his lips.
[D] He cried: „Bloody hell! It’s really close! We
should try not to move.‟
[I] He cried that it was really close and that they
should try not to move.
[D] He mouthed: „Bloody hell! It’s really close! We
should try not to move.‟
[I] He mouthed that it was really close and that they
should try not to move.
Katy was reading in the library and had put ear plugs
in to block out the other students studying nearby. She
absentmindedly began to click her tongue, not
realising that other people could hear her until a
concerned librarian tapped her lightly on the shoulder.
Page | 171
5
6
7
8
9
The nursery was filled with lots of young children,
screaming for attention from the busy childminder.
One boy in particular had been told off many times for
behaving disgracefully all day.
The three siblings knew they had to flee the city that
night if they were going to escape persecution and
crept along the dark lane. They climbed cautiously
into a truck and the eldest made sure her youngest
brother was completely hidden behind the wooden
crates.
[D] With her face turning red, she stressed to him:
„Listen. You have to be really good and not make a
sound for the next few hours.‟
[I] With her face turning red, she stressed to him that
he had to be really good and not make a sound for
the next few hours.
Harry dodged the flaming dragon, hearing it roaring
over the game's soundtrack, and reached the next level
of the computer game. Harry whooped euphorically
over the sound of his friends shouting out
congratulations.
[D] With her face white with worry, she lowered her
voice: „Listen. You have to be really good and not
make a sound for the next few hours.‟
[I] With her face white with worry, she lowered her
voice saying that he had to be really good and not
make a sound for the next few hours.
The class were supposed to be revising independently
but all the boys were passing around a new handheld
computer game under the teacher‟s nose. Harry almost
whooped when he managed to complete it and he leant
over to show his friends secretly below the table.
[D] Grinning from ear to ear, he boasted smugly: „I
told you that I’d manage it and I beat all of your
scores.‟
[I] Grinning from ear to ear, he boasted smugly that he
had told them he would manage it and he had beaten
all of their scores.
John and Fiona had been dancing near the stage for
hours and when the heavy metal concert ended, their
ears were ringing. They made their way to the VIP
area, hoping to catch a glimpse of the band, but were
stopped by a stern steward guarding the area.
[D] Winking, he said under his breath: „I told you that
I’d manage it and I beat all of your scores.‟
[D] Folding his arms across his chest, he boomed:
„I’m sorry but I need to see your tickets.‟
[I] Folding his arms across his chest, he boomed that
he was sorry but he needed to see their tickets.
The nightclub was very busy when an irritated drunk
man tried to punch a member of the bar staff. As the
police were called, two burly bouncers approached
him menacingly from both directions.
[D] He explained with an apologetic shrug: „I’m sorry
but I need to see your tickets.‟
[I] He explained with an apologetic shrug that he was
sorry but he needed to see their tickets.
The movie premiere was going well with the audience
sitting in tense anticipation as the thriller reached a
disquieting stage. Two ushers appeared noiselessly in
the aisle beside Sam and Fiona and motioned to a man
illegally recording the new film.
[D] Brusquely blocking his way, they said forcefully:
„Excuse me sir. We’d like you to accompany us
upstairs to the office.‟
[I] Brusquely blocking his way, they said forcefully
that they would like him to accompany them upstairs
to the office.
The raucous cheers rang out into the stadium, the fans
exploding with joy as the ball was headed into the
back of the net by the striker. Steven and his friends
were hugging each other, stomping and cheering.
[D] They leant towards the man and said in a hushed
voice: „Excuse me sir. We’d like you to accompany us
upstairs to the office.‟
[I] They leant towards the man and said in a hushed
voice that they would like him to accompany them
upstairs to the office.
It had been a tense Wimbledon match ending in a tie
and the game would be decided by sudden death. As
they waited for the UK duo to serve, the spectators
were warned not to distract the players.
[D] Dancing in his seat, he was cheering joyfully:
„Come on. If we win this match, we might have a shot
at the trophy.‟
[I] Dancing in his seat, he was cheering joyfully that if
they won the match, they might have a shot at the
trophy.
[D] Steven leaned in excitedly to his friend and
murmured: „Come on. If we win this match, we might
have a shot at the trophy.‟
[I] Steven leaned in excitedly to his friend and
murmured that if they won the match, they might
have a shot at the trophy.
[I] Winking, he said under his breath that he had told
them that he would manage it and he had beaten all
of their scores.
The carriage was virtually deserted at 11pm and John
and Fiona felt themselves falling asleep with the
motion of the train. They woke up groggily to see the
inspector standing over them, gently prodding their
shoulders.
Page | 172
10
11
12
13
14
The bars had just closed at 1am and the roads were full
of people merrily making their way home. As Helen
was walking in the other direction, Tara suddenly
remembered something and tried to catch her
attention.
Helen and Tara entered their flat, making sure not to
slam the door and wake up their flatmates. Helen
removed her heels and tip-toed across the hall when
Tara tapped her lightly on the shoulder.
[D] She called to her: „Oi! Don‟t forget! Paul’s
birthday party is on Friday so we’ll need to go
shopping.‟
[I] She called to her that Paul’s birthday party was on
Friday and they would need to go shopping.
[D] She discreetly said: „Oi! Don‟t forget! We‟ve got
Paul’s birthday party on Friday so we’ll need to go
shopping.‟
[I] She discreetly said that they would have Paul’s
birthday party on Friday so they would need to go
shopping.
As usual, the office was vacant and the two night
watchmen glanced at the CCTV screens sleepily from
time to time. Both men gulped as they felt a cold hard
blade pressed to their throats and soft footsteps on the
floor behind them.
It was the first day back at school and all the children
were very excited to see each other again. The journey
home was chaos for the poor bus driver with pupils
standing on their seats and shouting to friends.
[D] Finally the driver had enough, pulled over and
exclaimed angrily: „Listen. I want you to stay in your
seat and keep quiet if you want to get home tonight.‟
[I] Finally the driver had enough, pulled over and
exclaimed angrily that he wanted them to stay in their
seats and keep quiet if they wanted to get home
tonight.
[D] A ghostly voice breathed: „Listen. I want you to
stay in your seat and keep quiet if you want to get
home tonight.‟
The salsa teacher threw them straight into dancing,
barking out orders and dragging them to the front of
the room. Susan and Jane were partnering each other
and having a great time twirling to the fast-paced Latin
beats.
Susan and Jane were growing restless as the meeting
dragged on past five o‟clock and kept shooting sneaky
looks at each other. As their manager turned his back,
Susan rolled her eyes and leaned over to Jane.
[D] Giggling at her dizzy head, Susan shrieked: „Oh
dear. I think I’m going to need a drink after this.‟
[I] Giggling at her dizzy head, Susan shrieked that she
thought she was going to need a drink after this.
All that Gary and Shona could hear was the whistling
of the wind and the rain making hard thuds as it hit
their jackets. They held onto their hoods as they
struggled against the fierce gusts and tried to find the
way back to their hotel in the town.
[D] She said covertly: „Oh dear. I think I’m going to
need a drink after this.‟
[I] She said covertly that she thought she was going
to need a drink after this.
Gary and Shona took a late night stroll in a foreign
city, carrying a large amount of money. When they
took a wrong turn, Gary was concerned about being
mugged and was conscious of the few people around
them.
[D] Gary couldn‟t stand it any longer and roared above
the wind: „I should go and ask for directions from
someone in that restaurant.‟
[I] Gary couldn‟t stand it any longer and roared above
the wind that he should go and ask for directions
from someone in that restaurant.
The shop floor‟s speakers thudded with a new drum
and bass track as Laura looked through the old records
and found an early copy of her favourite album. She
didn't want to risk leaving it in the shop when her
mobile rang so she put her hand over her ear as she
answered.
[D] Gary moved closer so as not to be overheard and
said in Shona's ear: „I should go and ask for
directions from someone in that restaurant.‟
[I] Gary moved closer so as not to be overheard and
said in Shona's ear that he should go and ask for
directions from someone in that restaurant.
Laura returned to collect her developing slides and
found the lab full of other scientists working
independently. Her mobile began to vibrate, raising
eyebrows from her co-workers and she grimaced
apologetically.
[D] She said loudly into the mouthpiece: „I’m a little
bit busy at the moment but I’ll call you back in ten
minutes.‟
[I] She said loudly into the mouthpiece that she was a
little bit busy at the moment but she would call them
back in ten minutes.
[D] She quietly answered: „I’m a little bit busy at the
moment but I’ll call you back in ten minutes.‟
[I] A ghostly voice breathed that he wanted them to
stay in their seats and keep quiet if they wanted to get
home tonight.
[I] She quietly answered that she was a little bit busy
at the moment but she would call them back in ten
minutes.
Page | 173
15
16
17
18
The thunderous drone of the boat's motor added to the
splashing of the waves and the wind in their ears as
Ian and David made their way out of the harbour.
Steering the boat, Ian could just make out David
saying something about sandwiches below deck.
Ian and David liked going to the cinema but resented
paying for the overpriced refreshments so they had
sneaked in their own. David was rustling through his
bag as the film began to play and the room fell silent.
[D] Ian tried to see into the hull and called down: „If
you’re looking for the crisps and sandwiches, they’re
in the red rucksack.‟
[I] Ian tried to see into the hull and called down that if
he was looking for the crisps and sandwiches, they
were in the red rucksack.
Leanne‟s son had promised her that he would tidy his
room when he got home from school. She was furious
to find him watching a film, his speakers blaring and
his room still a mess.
[D] Ian leaned over and whispered: „If you’re looking
for the crisps and sandwiches, they’re in the red
rucksack.‟
[I] Ian leaned over and whispered that if he was
looking for the crisps and sandwiches, they were in
the red rucksack.
On Sunday evening, Leanne saw her son reading
comic books in his room which was still messy despite
his promise. As she spoke to him in his room, she was
furious but tried not to disturb her youngest daughter
sleeping next door.
[D] She lunged to turn the TV off and exploded:
„That‟s it! You’re going to be grounded until this
room is sparkling.‟
[I] She lunged to turn the TV off and exploded that he
was going to be grounded until his room was
sparkling.
The gym was full of people training, running, kicking
and punching to heavy bass pumping from the ceiling
speakers. Helen had been trying to improve her
kickboxing for weeks and her trainer was pushing her
to the limit.
[D] She pulled his ear and hissed: „That‟s it! You’re
going to be grounded until this room is sparkling.‟
[D] He yelled at her: „Come on. I know you can do
better than this so I want you to repeat the exercise.‟
[D] He said in a reassuring voice: „Come on. I know
you can do better than this so I want you to repeat
the exercise.‟
[I] He said in a reassuring voice that he knew that she
could do better than that so he wanted her to repeat
the exercise.
The library was full of students studying seriously for
their final exams. Patrick looked at his watch and
realised that he and Annie should leave for their exam.
[I] He yelled at her that he knew that she could do
better than that so he wanted her to repeat the
exercise.
In her bedroom, Annie put on some cheerful music
while she was packing her suitcase for her holiday to
Spain. Downstairs, Patrick looked at his watch and
realised that they would have to leave soon to catch
their flight.
[D] He stood at the bottom of the stairs and shouted:
„Come on, darling. We should leave in about half an
hour if we don’t want to be late.‟
19
[I] He stood at the bottom of the stairs and shouted
that they should leave in about half an hour if they
didn’t want to be late.
Daniel was racing his new motor wildly in the street
and he lost control of the vehicle, rushing into a lorry.
In the hospital, he knew his father would be incredibly
angry with him for speeding.
[D] When his father arrived, he slammed his fist into
the hospital door and roared: „Jesus Christ! You're
damn lucky to be alive today.‟
[I] When his father arrived, he slammed his fist into
the hospital door and roared that he was damn lucky
to be alive today.
[I] She pulled his ear and hissed that he was going to
be grounded until his room was sparkling.
The teacher set the class to work silently on their
individual projects and began to mark their essays. He
was shocked at the unexpectedly low standard of
Helen's work and asked her to speak to him outside the
class to stop the other pupils hearing.
[D] Trying not to disturb the other students nearby, he
whispered to Annie: „Come on, darling. We should
leave in about half an hour if we don’t want to be
late.‟
[I] Trying not to disturb the other students nearby, he
whispered to Annie that they should leave in about
half an hour if they didn’t want to be late.
Daniel had been very lucky to survive the car accident
when a drunk driver hit his car. His father looked
heartbroken when he saw his son in the hospital bed
for the first time.
[D] His face was white as he said in a sad voice: „Jesus
Christ! You're damn lucky to be alive today.‟
[I] His face was white as he said in a sad voice that he
was damn lucky to be alive today.
Page | 174
20
21
22
23
Alex was working in a busy antiques store on the local
high street. At the most hectic point of the morning, an
old man asked her repeatedly about one of their tables
but wasn‟t listening carefully to her response.
The two art dealers were searching through the small,
dusty antiques shop when one of them spotted a
bargain. He didn‟t want to draw the owner‟s attention
to the true rarity and expense of the piece.
[D] She couldn‟t make herself any clearer as she
exclaimed: „Listen. The table is £2500 and is from the
late 18th century.‟
[I] She couldn‟t make herself any clearer as she
exclaimed that the table was £2500 and was from the
late 18th century.
The Masterchef studio kitchen was filled with the
clanging of pots and pans and the sizzling of food. The
loudspeaker was activated in the hectic kitchen for the
presenter to make an announcement to the three
finalists.
[D] He turned to his brother and gestured, discreetly
saying: „Listen. The table is £2500 and is from the
late 18th century.‟
[I] He turned to his brother and gestured, discreetly
saying that the table was £2500 and was from the late
18th century.
Heather had missed her class exam and had to sit the
test alone under the supervision of her English teacher.
As the final half an hour approached, she was focused
entirely on her work and the teacher kindly reminded
her of the time.
[D] He tapped the microphone and declared with
authority: „OK. It’s quarter past one so there’s half
an hour left until the interval.‟
[I] He tapped the microphone and declared with
authority that it was quarter past one so there was
half an hour left until the interval.
James was annoyed when the documents he had asked
to be sent were returned to the company in the post.
He was in a foul mood and stormed through to his
assistant‟s office, not bothering to knock on the door.
[D] He cleared his throat, gently saying: „OK. It’s
quarter past one so there’s half an hour left until the
interval.‟
[I] He cleared his throat, gently saying that it was
quarter past one so there was half an hour left until
the interval.
The woods were still and motionless as the car pulled
into the lane and turned off its engine. A knock on the
glass made James start and he rolled down his window
but couldn't see the agent‟s face.
[D] Banging the package on the desk, James angrily
erupted: „Make sure this gets to the right person.‟
[I] Banging the package on the desk, James angrily
erupted that he should make sure that it got to the
right person.
Erica had taken her daughter to an indoor adventure
play area where there was an assault course, a ball
pool and inflatable toys. After an hour of watching her
play with the other children, she motioned
energetically in the crowd to get her attention.
[D] Handing him the package, James muttered: „Make
sure this gets to the right person.‟
[I] Handing him the package, James muttered that he
should make sure that it got to the right person.
[D] Erica shouted out: „I’m afraid that’s time up. I’ll
meet you at reception.‟
[D] She placed her hand lightly on her shoulder,
saying soothingly: „I’m afraid that’s time up. I’ll
meet you at reception.‟
[I] She placed her hand lightly on her shoulder, saying
soothingly that she was afraid that it was time up and
that she would meet her at reception.
An ambulance siren could be heard in the distance as
Kelly edged past distressed neighbours to see the
accident. She saw the man‟s body lying in the road,
surrounded by the eerily motionless crowd.
[I] Erica shouted out that she was afraid that it was
time up and that she would meet her at reception.
24
Everyone in Kelly‟s living room was intrigued as the
documentary showed the plastic surgery taking place.
However, as the camera zoomed in on the gory scene,
the living room exploded with utter disgust.
[D] Kelly couldn‟t believe what she was seeing and
yelled: „Oh my god. I’ve never seen anything so
disturbing.‟
[I] Kelly couldn‟t believe what she was seeing and
yelled that she had never seen anything so disturbing.
Erica loved going for an aromatherapy massage and
felt her muscles soothe as the hour went on. She felt
herself drifting off with the smell of the oils and the
relaxing music when the masseuse turned down the
CD.
[D] Kelly covered her mouth in shock and murmured:
„Oh my god. I’ve never seen anything so disturbing‟.
[I] Kelly covered her mouth in shock and murmured
that she had never seen anything so disturbing.
Page | 175
Appendix 4 speaking rate stimuli
[D] = direct speech; [I] = indirect speech; [Q] = comprehension question (for Experiment 10 only)
The bold and italic texts in direct and indirect speech quotations are the critical regions.
Item
1
2
Fast-speaking story
It was 11am in the morning when the fire alarm went
off. Hearing people running down the corridors, Mary
grabbed her jacket and burst into Peter‟s office next
door.
Slow-speaking story
It was 11am in the morning when the fire alarm went
off. Knowing that this was just a test, Mary put on her
jacket and walked into Peter‟s office next door.
[D] She shouted: „Peter, quick, we have to leave
immediately because the building is on fire!‟
[I] She urged Peter to leave immediately because the
building was on fire.
[D] Tongue-in-cheek, she said: „Peter, quick, we have
to leave immediately because the building is on fire!‟
[I] Tongue-in-cheek, she told Peter to leave
immediately because the building was on fire.
Confused, Peter replied: „Wasn‟t there an e-mail about
a fire alarm test this morning?‟
Peter just replied: „Very funny – I‟ve seen the e-mail
about the fire alarm test as well!‟
[Q] Did Mary think it was a real fire?
It was a Sunday morning during the summer vacation.
At the airport, Cheryl and Gareth were boarding the
flight to Paris for their honeymoon and there were only
5 minutes left before the plane was taking off. Cheryl
started panicking because she couldn‟t find her
passport.
[Q] Did Mary think it was a real fire?
It was a Sunday morning during the summer vacation.
Cheryl and Gareth finally found the time to sort some
personal documents and other clutter that had been
piling up on the kitchen table. Cheryl was looking for
her passport and became confused.
[D] Hastily, she said to Gareth: „Oh no! I am pretty
sure it was in that sheath, but I cannot find it
anymore!‟
[I] Hastily, she said to Gareth that she was pretty sure
it was in that sheath, but she couldn’t find it anymore.
[D] She turned to Gareth and said: „Hmm, that‟s weird.
I am pretty sure it was in that sheath, but I cannot
find it anymore!‟
[I] She turned to Gareth and said that it was weird
because she was pretty sure it was in that sheath, but
she couldn’t find it anymore.
Gareth went straight back to the duty-free shop, and
luckily, he found the passport by the till.
Gareth looked under a pile of bills, and luckily, he
found her passport there.
[Q] Did they manage to catch the flight?
3
It was a typical British day, rainy and gloomy. Sixteen
year-old pianist Bobby was going to play in the
quarter-final of a local talent competition. He was
extremely nervous before his performance.
[Q] Did Cheryl need her passport at the time?
It was a typical British day, rainy and gloomy. At
Glasgow Royal Infirmary, an old man was dying, and
too weak to sit up. His family members were sitting
around the bed, feeling sad. He wanted to say
something, so his daughter placed a cushion under his
head.
[D] His mother encouraged him but he was all shaking
and said: „No! I can‟t do it! This is the end of the
journey because it is unlikely that I will make it this
time.‟
[I] His mother encouraged him but he was all shaking
and said that he couldn‟t do it and that it was the end
of the journey because it was unlikely that he would
make it this time.
[D] Slowly, he looked around and said: „I‟m grateful
you‟re all here. This is the end of the journey because
it is unlikely that I will make it this time.‟
[I] Slowly, he looked around and said that he was
grateful for their coming and that it was the end of the
journey because it was unlikely that he would make it
this time.
His mother tried to calm him down, saying that it‟s not
the winning that counts, but the taking part.
Then he closed his eyes and everyone burst into tears.
[Q] Was Connie anxious about her coming
performance in the quarter-final?
[Q] Was the man too weak to open his mouth?
Page | 176
4
At the beach, William and Phillip were enjoying the
sunshine and a fresh sea breeze. Suddenly, William
noticed a piece of wood floating on the water with
something lying on it that looked like an unconscious
person.
At the beach, William and Phillip were enjoying the
sunshine and a fresh sea breeze. They were discussing
the hypothetical case of what to do if they saw someone
drowning.
[D] Without hesitation, he exclaimed: „Phil, quick! We
need to call for help because there might be someone
in trouble!‟
[D] William thought there would be little one could do,
but Phil objected: „Look. The most important thing is to
call for help because there might be someone in
trouble.‟
[I] William thought there would be little one could do,
but Phil objected that the most important thing would be
to call for help because there might be someone in
trouble.
[I] Without hesitation, he exclaimed that they needed to
call for help because there might be someone in
trouble.
5
Phillip immediately grasped his mobile phone and
dialled 999.
After a while, the conversation moved on to a different
topic.
[Q] Was there a person in danger?
It was at that peaceful house in Dumbarton where Mrs.
Harris was preparing lunch for her children. Suddenly,
the phone rang. It was Mr. Harris‟s business partner,
urgently informing Mrs. Harris that her husband has had
a car crash this morning.
[Q] Was there a person in danger?
It was at that peaceful house in Bristol where PCs Bryce
and Rawlinson had some sad news to impart. The owner
of the house, Mr. Harris, was killed in a car crash earlier
this morning. Paramedics had rushed on the scene in
time, but they couldn‟t save him.
[D] He spoke hastily: „Mrs. Harris, please come
immediately! Your husband was involved in a collision
with a heavy goods vehicle and is now in hospital!‟
[D] With sorrow, Bryce told Mrs. Harris: „Mrs. Harris, I
am afraid that your husband was involved in a collision
with a heavy goods vehicle and he is now in hospital.‟
[I] With sorrow, Bryce told Mrs. Harris that her
husband was involved in a collision with a heavy goods
vehicle and was now in hospital.
[I] He hastily told Mrs. Harris to come immediately
because Mr. Harris was involved in a collision with a
heavy goods vehicle and was now in hospital.
Mrs. Harris was shocked, and headed for the hospital
right away.
Mrs. Harris sobbed. Apparently, she could not believe
it.
[Q] Was Mr. Harris dead?
6
[Q] Was Mr. Harris dead?
On Valentine‟s Day, Simon at long last worked up the
courage to propose to Emma. Yet to his horror, she
refused. He felt humiliated and blamed Emma for her
brusque reply.
On Valentine‟s Day, Simon at long last worked up the
courage to propose to Emma. Yet to his horror, she
refused.
[D] Emma fought back: „Come on, Simon! How could
you blame me for this? I am simply not ready for such
a commitment!‟
[I] Emma fought back, saying that Simon could not
really blame her for that and that she was simply not
ready for such a commitment.
[D] He asked for the reason and Emma replied calmly:
„You are a really nice bloke, Simon, but I am simply
not ready for such a commitment.‟
[I] He asked for the reason and Emma replied calmly
that although she thought he was a really nice bloke,
she was simply not ready for such a commitment.
Simon was speechless, realising that he was humiliating
himself.
Simon was upset for a moment but determined not to
give up.
[Q] Was Simon‟s proposal successful?
[Q] Was Simon‟s proposal successful?
Page | 177
7
8
9
Alice was looking for pheasant paintings for her art
history dissertation. Today her boyfriend John bumped
into a weird-looking man with greasy hair and thick
glasses who, amazingly, claimed to possess a sizeable
collection of what Alice was looking for.
Alice was selling paintings in a small shop downtown.
One day, a weird-looking man with greasy hair and
thick glasses came into her shop.
[D] John was excited and phoned Alice immediately:
„Alice, you wouldn‟t believe it!.. I just met one of the
world’s keenest collectors of pheasant paintings!‟
[I] John was excited and phoned Alice immediately,
saying that he just met one of the world’s keenest
collectors of pheasant paintings.
[D] He looked around and said nonchalantly: „You may
be surprised to learn that I am one of the world’s
keenest collectors of pheasant paintings.‟
[I] He looked around and said nonchalantly that she
might be surprised to learn that he was one of the
world’s keenest collectors of pheasant paintings.
„Really?!‟ Alice couldn‟t believe her ears.
Alice replied: „Uhm.. I don‟t think we have any at the
moment.‟
[Q] Was the man interested in pheasant paintings?
On the train to London, Olivia was very nervous about
her upcoming job interview. Looking through the
window, she reached for a cigarette without thinking,
just to calm herself down.
[Q] Was the man interested in pheasant paintings?
Olivia and her 5 year old daughter were on a train to
London. Looking through the window, a thoughtless
youth lit a cigarette and was immediately reprimanded
by other passengers. Olivia‟s daughter was curious and
asked why one cannot smoke on the train.
[D] Instantly, a middle-aged man next to her
complained: „Oi! Stop it! Smoking on public transport
is illegal!‟
[I] Instantly, a middle-aged man next to her complained
that she should stop that because smoking on public
transport was illegal.
[D] Olivia smiled and answered: „Well, since it may
harm other passengers‟ health, smoking on public
transport is illegal.‟
[I] Olivia smiled and answered that since it may harm
other passengers‟ health, smoking on public transport
is illegal.
Olivia put the cigarette back into the box and felt even
more stressed after this incident.
The girl thought for a second and said: „Why can‟t they
just hold their breath?‟
[Q] Did Olivia smoke on the train?
PhD student Ella was summoned to her supervisor Jim‟s
office to give a report on her current progress. Ella
asked for an extension but Jim looked rather impatient.
[Q] Did Olivia answer her daughter patiently?
PhD student Ella was summoned to her supervisor Jim‟s
office to give a report on her current progress. Ella
asked for an extension but Jim seemed to have a better
idea.
[D] Hastily, he said: „No way! We really need to have
those data in by next month for that conference!‟
[I] Hastily, he said that they really needed to have those
data in by next month for that conference.
[D] He said: „Actually…, it might be good to have
those data in by next month for that conference.‟
[I] He said that it might be good to have those data in
by next month for that conference.
Ella was gutted because it meant that she would have to
work like mad over the following weeks.
Ella understood his point and decided to work even
harder over the following weeks.
[Q] Was Ella‟s request for an extension rejected
immediately?
[Q] Was Ella‟s request for an extension rejected
immediately?
Page | 178
10
11
In the board room, assistant manager Craig was giving a
presentation on how to improve the company‟s sales
figures. A fellow manager was far from impressed and
openly criticised Craig‟s public speaking skills.
In the board room, assistant manager Craig was giving a
presentation on how to improve the company‟s sales
figures. When he noticed that his colleagues looked far
from impressed, he felt embarrassed.
[D] Craig kind of lost his temper and was quick to
defend himself: „What do you expect?! I never had to
give this kind of presentation before, so I am not very
experienced!‟
[I] Craig kind of lost his temper and was quick to
defend himself, saying that he never had to give this
kind of presentation before, so he was not very
experienced.
[D] Craig hesitantly explained: „Sorry, I never had to
give this kind of presentation before, so I am not very
experienced.‟
Everyone else in the room went quiet.
Everyone else in the room went quiet.
[Q] Was Craig apologetic about his performance?
It was 5.30 pm and people were ready to leave the
office. Elaine seemed to be rather agitated, cursing a lot
in front of her screen. Steven went to Elaine‟s desk
asked and what‟s wrong.
[Q] Was Craig apologetic about his performance?
It was 5.30 pm and people were ready to leave the
office. Steven went to Elaine‟s desk and saw her
slumped in her chair.
[D] Elaine complained: „Just look at this! The amount
of admin is killing me at the moment, and there seems
to be no end to it in sight!‟
[D] He asked what‟s wrong, and Elaine replied with a
listless expression on her face: „I‟m totally worn out.
The amount of admin is killing me at the moment, and
there seems to be no end to it in sight.‟
[I] He asked what‟s wrong, and Elaine replied with a
listless expression on her face that the amount of admin
was killing her at the moment, and that there seemed
to be no end to it in sight.
[I] Elaine complained that the amount of admin was
killing her at the moment, and that there seemed to be
no end to it in sight.
12
[I] Craig hesitantly explained that he never had to give
this kind of presentation before, so he was not very
experienced.
Steven knew exactly what she was talking about.
Steven knew exactly what she was talking about.
[Q] Was Elaine agitated about admin?
It was half past seven and people were unusually busy
at the Aberdeen restaurant. The chef could hardly cope
with the pressure of serving a bus-load of tourists who
had just arrived. After a couple of complaints, the
restaurant manager went berserk.
[Q] Was Elaine agitated about admin?
It was half past seven and people were unusually busy
at the Aberdeen Cooking College. The students had just
finished their final exam and were awaiting the results.
In front of Mike‟s dishes, the examiners frowned.
[D] He dashed straight into the kitchen and shouted: „Is
this how to treat our customers!? The fish is burnt, the
vegetables overcooked, and the soup is actually stone
cold!‟
[D] After exchanging their views, the head examiner
finally announced a verdict: „Mike, we really appreciate
the efforts you‟ve made, but the fish is burnt, the
vegetables overcooked, and the soup is actually stone
cold.‟
[I] After exchanging their views, the head examiner
finally announced a verdict, saying that although they
really appreciated the efforts Mike had made, the fish
was burnt, the vegetables overcooked, and the soup
was actually stone cold.
[I] He dashed straight into the kitchen and shouted
whether this was how to treat their customers, pointing
out that the fish was burnt, the vegetables overcooked,
and the soup was actually stone cold.
Everything went instantly quiet in the kitchen, except
for the occasional bubbling noises from the cooker.
Mike felt upset because this could only mean that he
had failed.
[Q] Was the manager angry with the chef?
[Q] Did the examiners fail Mike immediately?
Page | 179
13
Jenny‟s fiancé Ben had a bad motorcycle accident
which was likely to leave him paraplegic for the rest of
his life. Calling from the hospital, he said farewell to
Jenny, anticipating that she might leave him because of
his condition.
Jenny and Ben were having a candle light dinner to
celebrate the first anniversary of their relationship. Yet
at some point, Jenny hesitantly mentioned that because
of her new job she had to move to another city.
[D] Jenny was crying and shouting: „No! I’m not
leaving you! I want to be with you whatever happens
because I love you!‟
[D] After a sip of champagne, Ben held her hands and
said very calmly: „Don‟t worry, darling. I’m not leaving
you. I want to be with you whatever happens because I
love you.‟
[I] After a sip of champagne, Ben held her hands and
said very calmly that he was not leaving her and that
he wanted to be with her whatever happens because he
loved her.
[I] Jenny was crying and shouting that she was not
leaving him and that she wanted to be with him
whatever happened because she loved him.
14
15
Jenny‟s reassurance came as a great relief.
Ben‟s reassurance came as a great relief.
[Q] Did Ben have a motorcycle accident?
Shauna was a young, talented actress and singer. She
was desperate for a part in the new musical project but
her boyfriend was ill and blamed her for not looking
after him.
[Q] Did Jenny have a new job?
Theatrical agent Peter was visited by one of his clients.
Shauna was a young, talented actress and singer. She
was desperate for a part in the new musical project.
[D] Shauna was angry with his attitude, blustering:
„Don‟t be ridiculous! I am doing this because without
this job, I can’t even pay for the electricity bill!‟
[I] Shauna was angry with his attitude, blustering that
she was doing this because without this job, she
couldn’t even pay for the electricity bill.
[D] Shauna hummed and hawed: „It is embarrassing to
say, but…, I am doing this because without this job, I
can’t even pay for the electricity bill!‟
[I] Shauna hummed and hawed with embarrassment,
saying that she was doing this because without this job,
she couldn’t even pay for the electricity bill.
Her boyfriend was speechless and realised how childish
he was.
Peter was moved and promised he would try his best to
help her.
[Q] Did Shauna think her boyfriend was sensible?
Harry and Leona were having a traditional Sunday
roast. Harry was constantly talking with his mouth full.
Suddenly, he stopped and started choking, with his face
turning bright red.
[Q] Did Shauna need money?
Red Cross nurse Leona was teaching the public how to
make an emergency call. She first introduced a list of
important information that should be provided.
[D] Leona was panicking and dialled 999: „Help! My
husband is choking on his food and I actually don’t
know what to do! It looks rather serious!‟
[D] Then she demonstrated how NOT to do it by
mocking a nonchalant woman saying: „It seems as if my
husband is choking on his food and I actually don’t
know what to do. It looks rather serious…‟
[I] Then she demonstrated how NOT to do it by
mocking a nonchalant woman saying that it seemed as if
her husband was choking on his food and that she
actually didn’t know what to do, adding that it looked
rather serious.
[I] Leona was panicking and dialled 999, saying that
her husband was choking on his food and that she
actually didn’t know what to do, adding that it looked
rather serious.
The person at the other end instantly gave her some
emergency instructions.
Some people in the audience started to chuckle.
[Q] Was Harry having a soup?
[Q] Was Leona a school teacher?
Page | 180
16
17
18
It was the last day of John Lewis‟s summer sale when
two girls were browsing around. Helen was very fond
of a black skirt and shrieked when she found it was
heavily discounted.
It was the last day of John Lewis‟s summer sale when
two girls were browsing around. Helen felt almost
heartbroken when she found that her favourite skirt was
heavily discounted – she and her friend Abbey were
completely skint at the moment.
[D] She could not wait to show it to her friend and
shouted: „Hey Abbey! Look at this one! Gorgeous isn‟t
it!? It is exactly what I always wanted, and now it is
even more than 50% off!‟
[I] She could not wait to show it to her friend Abbey.
Excited, she asked her to have a look, shouting that it
was exactly what she always wanted, and now it was
even more than 50% off.
[D] With a deep-drawn sigh, she said: „Look, Abbey.
Gorgeous isn‟t it… It is exactly what I always wanted,
and now it is even more than 50% off…‟
Abbey agreed: „I‟m sure it will look great on you.‟
Abbey tried to comfort her: „I‟m sure there will be other
good offers next time round.‟
[Q] Was Helen excited about the skirt?
It was 1.55pm when the Riley‟s were about to head off
to the train station to collect granddad.
[Q] Was Helen fond of the skirt?
It was 1.55pm when the Riley‟s were about to head off
to the train station to collect granddad. Fourteen yearold Karen was asking why they went so early, given
that the train was due to arrive at 3pm.
[D] Mrs Riley was panicking because she couldn‟t find
the car keys: „Jesus! We‟ve got only five minutes to go!
If we don’t arrive in time, granddad might easily get
lost, considering how forgetful he sometimes is!‟
[I] Mrs Riley was panicking because she couldn‟t find
the car keys. She realized they‟ve got only five minutes
to go, and said that if they didn’t arrive in time,
granddad might easily get lost, considering how
forgetful he sometimes was.
[D] Her mum replied with a smile: „Well, if we don’t
arrive in time, granddad might easily get lost,
considering how forgetful he sometimes is.‟
„Don‟t panic,‟ her husband said, „I‟ve already called a
cab.‟
Karen remained defiant: „But it is Sunday, and there is
almost no traffic!‟
[Q] Will the train arrive at 2pm?
At Glasgow‟s Royal Infirmary, surgeons and nurses
were trying their best to save a man‟s life. The minutes
kept ticking away and the head surgeon looked
concerned.
[Q] Will the train arrive at 2pm?
At Glasgow‟s Royal Infirmary, the students were shown
some important medical emergency procedures.
[D] He was shouting: „Quick! Set up the life-support
machine and stop the bleeding!‟
[D] The tutor gave precise instructions: „When the
patient comes in, first set up the life-support machine
and stop the bleeding!‟
[I] The tutor gave precise instructions, saying that when
the patient comes in, first they should set up the lifesupport machine and stop the bleeding.
[I] He ordered to quickly set up the life-support
machine and stop the bleeding.
[I] With a deep-drawn sigh, she said to Abbey that it
was exactly what she always wanted, and now it was
even more than 50% off.
[I] Her mum replied with a smile, saying that if they
didn’t arrive in time, granddad might easily get lost,
considering how forgetful he sometimes was.
Nurses followed his instructions promptly but the man
showed little sign of life.
The students were taking notes and listened carefully.
[Q] Was it important to set up the life-support machine?
[Q] Was it important to set up the life-support machine?
Page | 181
19
20
21
It was Tuesday, 6th June 1944. At Omaha Beach, allied
soldiers braved artillery fire, mortar attacks and allround danger. Enemy gunfire poured at the approaching
allies and kept them at mid-distance.
At Omaha Beach, allied soldiers braved artillery fire,
mortar attacks and all-round danger. In the days before
the invasion, the assault had been meticulously planned
at allied headquarters.
[D] The captain of the 1st Division spurred his soldiers
on: „There is not much time! We must silence those
machine guns before the reinforcements arrive!‟
[I] The captain of the 1st Division spurred his soldiers
on, shouting that there was not much time, and that they
must silence those machine guns before the
reinforcements arrive.
[D] Back then, the officer in charge explained: „We
must silence those machine guns before the
reinforcements arrive.‟
[I] Back then, the officer in charge explained that they
must silence those machine guns before the
reinforcements arrive.
Then he jumped up and charged forward.
Obviously, the success of the whole operation depended
on that.
[Q] Should the machine guns be silenced?
A heavily armed robbery took place earlier this
morning. Within five minutes, five police cars showed
up at the scene, but the busy streets made it difficult to
locate the getaway culprits.
[Q] Should the machine guns be silenced?
The West End festival was coming soon and the
organisers were discussing details of the live television
coverage. John considered setting up several cameras,
but the busy streets made it difficult to locate the
parade.
[D] In the heat of the situation, the chief made a prompt
decision and ordered: „Block the entire West End! We
need a helicopter to cover the whole area!‟
[I] In the heat of the situation, the chief made a prompt
decision and ordered to block the entire West End,
adding that they needed a helicopter to cover the whole
area.
[D] Pointing at the map, John said: „Hmm, not sure
whether street cameras would work. Perhaps we need a
helicopter to cover the whole area.‟
[I] Pointing at the map, John said that he was not sure
whether street cameras would work, adding that perhaps
they needed a helicopter to cover the whole area.
Then he jumped into his car and headed for the nearest
exit.
Unfortunately, this plan was well out of budget.
[Q] Was a helicopter needed?
Throughout human history, wars have caused many
deaths. Today, United Nations Peacekeeping Forces
were ambushed by Taliban forces in northern
Afghanistan. They returned fire immediately, but
suffered heavy losses.
[Q] Was a helicopter needed?
Throughout human history, wars have caused many
deaths. Recently, Tom was interviewing a 75 year-old
Vietnam veteran. When Tom asked about the situation
in the Battle of Hue, the old soldier paused for a while
to recall his memories.
[D] The officer in charge hastily called for
reinforcements, shouting: „We are under attack, facing
mortar fire and rocket propelled grenades!‟
[I] The officer in charge hastily called for
reinforcements, shouting that they were under attack,
facing mortar fire and rocket propelled grenades.
[D] Then he slowly replied: „On the night of January
30, we were under attack, facing mortar fire and
rocket propelled grenades..‟
[I] Then he slowly replied that on the night of January
30, they were under attack, facing mortar fire and
rocket propelled grenades.
As he finished his message, a shell exploded beside
him.
His facial expression gave a clue of the terror he felt at
the time.
[Q] Was the story about the Vietnam war?
[Q] Was the story about the Afghanistan war?
Page | 182
22
23
It was almost midnight and moonlight dripped through
the leaves on the pavements. In the workshop, the cruel
overseer was driving his workers on, as he was under
pressure to meet his targets.
It was almost midnight and moonlight dripped through
the leaves on the pavements. Robert and Chloe just
came back from a party. Robert paid the cab driver and
wanted to go to sleep straight away. But Chloe was far
from tired and fancied another drink.
[D] When one of the workers started yawning, he
instantly shouted: „Oi, lazybones! The day has actually
only just begun!‟
[I] When one of the workers started yawning, he
instantly shouted at her, calling her lazybones, and that
the day had actually only just begun.
[D] She patted Robert‟s cheek and said: „Stop yawning,
my friend. The day has actually only just begun.‟
Although his methods may have been effective, he
certainly was not very popular.
Robert rolled his eyes, but agreed to have one last pint
somewhere.
[Q] Was the overseer under pressure?
One morning, six-year-old Sara was taking a lot of time
to get ready for school. Her dad got increasingly angry
with her because the school bus was due to arrive soon.
[Q] Did Chloe fancy another drink?
One morning, six-year-old Sara was taking a lot of time
to get ready. She was secretly hoping that she would
miss the school bus if she kept dragging on.
[D] He said: „Sara! If you are not ready within five
minutes, I’ll have to drive you again!‟
[D] Her dad could easily see what she was up to and
said in a sarcastic tone: „Just take your time.. If you are
not ready within five minutes, I’m happy to drive you
again.‟
[I] Her dad could easily see what she was up to and said
in a sarcastic tone that she could just take her time,
adding that if she was not ready within five minutes,
he’d be happy to drive her again.
[I] He said to Sara that if she was not ready within five
minutes, he’d have to drive her again.
24
[I] She patted Robert‟s cheek and told him to stop
yawning because the day had actually only just begun.
Reluctantly, Sara complied and put on her coat.
Reluctantly, Sara complied and put on her coat.
[Q] Was Sara taking a lot of time?
Millionaire Joseph was addicted to betting on horses.
Tipped by a so-called „insider‟, he recently placed an
enormous bet, but shockingly, the horse had lost.
[Q] Was Sara taking a lot of time?
Millionaire Joseph was addicted to betting on horses.
Tipped by a so-called „insider‟, he recently placed an
enormous bet, but shockingly, the horse had lost.
[D] Angry with his informant, Joseph shouted furiously
on the phone: „Where did your bloody information
come from!? That was a huge amount of money –
almost one million pounds!‟
[I] Angry with his informant, Joseph shouted furiously
on the phone, asking where the information had come
from, because that was a huge amount of money –
almost one million pounds.
[D] Realising that he was nearly bankrupt, Joseph
slumped back into his chair and started talking to
himself: „That was a huge amount of money – almost
one million pounds…‟
[I] Realising that he was nearly bankrupt, Joseph
slumped back into his chair and started talking to
himself, repeatedly saying that it was a huge amount of
money – almost one million pounds.
Then he realised that it was his wife speaking at the
other end.
Suicidal thoughts were creeping up his mind.
[Q] Did Joseph bet a small amount of money?
[Q] Did Joseph bet a small amount of money?
Page | 183
Appendix 5 Fixation Density Disparity
Fixation Density Disparity (FDD) is designed to quantify the fluency of the initial reading
of a relatively long critical region (e.g., a sentence) in reading experiments using eyetracking. The measure primarily captures the spatial distribution of fixations in a given
region. As such, it is independent of the reading-time measures of interest – thus, it is
unlikely to introduce a bias when used as an outlier criterion for reading-time analyses.
0 (if f = 0),
During initial reading of a region, FDD =
F
(if f > 0)
f
F = the number of surrounding fixations
where
f = the number of surrounded fixations
A given fixation is characterised as a surrounded fixation when there are other fixations
within the vicinity of 5 characters to the left or right of that fixation (blue circled fixation
in Example 1 below). These neighbouring fixations are surrounding fixations (fixations
indicated by purple arrows in Example 1). The window of 5 characters to the left and right
of a given fixation is motivated by a commonly reported average saccade size of 5-7
characters between fixations during normal reading (e.g., Morrison & Rayner, 1981)
A high FDD value indicates that there are clusters of spatially densely distributed fixations
during initial reading of the critical region (e.g., due to frequent re-inspections of words),
suggesting faltering reading (Example 1). By contrast, a low FDD value occurs when
fixations are evenly distributed across the critical region, indicating reasonably fluent
reading; an FDD of 0 suggests perfectly fluent reading (Example 2).
Note. A cut-off FDD value of 2 (see outlier definition in paper) implies that the average
go-past fixation per region may be surrounded by no more than two fixations within its 5character vicinity.
Page | 184
Example 1: Faltering reading (each green square represents a fixation; the lines between
them represent saccades; each grey column represents a character position)
Fixation ID
Number of surrounding
fixations at each given
fixation
Surrounded by other
fixations? (0=no, 1=yes)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
total
1
1
3
6
5
3
4
3
4
1
1
0
1
1
1
1
F =36
FDD
2.40
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
f = 15
Example 2: Fluent reading.
Fixation ID
Number of surrounding fixations at each given fixation
1
2
3
4
5
total
0
0
0
0
0
F=0
Surrounded by other fixations? (0=no, 1=yes)
0
0
0
0
0
f=0
FDD
0
Reference
Morrison, R. E., & Rayner, K. (1981). Saccade size in reading depends upon character
spaces and not visual angle. Perception and Psychophysics, 30(4), 395-396.
Page | 185
Appendix 6 Supplementary measures (Experiment 10)
The go-past time measures reported in the paper are conceptually equivalent to the product
of (i) the numbers of go-past fixations launched on the critical regions and (ii) the average
durations of those go-past fixations. To examine whether the reported reading rate
modulations in Experiment 10 were due to either (i) or (ii) (or indeed a combination of
both), we conducted further analyses on these sub-measures.
(i)
Numbers of fixations
Tables I and II below show the average numbers of go-past (respectively no-regressions
go-past) fixations launched in each condition of Experiment 10. There was a significant
Context  Quoting Style interaction in the number of go-past fixations launched (F1(1, 47)
= 18.44, p < .001; F2(1, 23) = 6.24, p = .02) and, at least marginally, in the number of noregressions go-past fixations launched (F1(1, 47) = 10.15, p < .002; F2(1, 23) = 4.01, p
< .06).
Table I. Numbers of go-past fixations launched in Experiment 10 (with standard deviations in brackets).
Also shown are results from 2-tailed paired sample t-tests examining the simple effects of Context and
Quoting Style.
Direct
Speech
Fast-speaking Context
Slow-speaking Context
t(47)
By
Subjects
P
Context
Contrast
t(23)
By
Items
P
9.5 (3.6)
10.1 (3.7)
-2.86
.006
-1.78
.09
Quoting Style Contrast
By Subjects
By Items
t(47)
p
t(23)
p
10.3 (4.1) -2.72
.01 -2.54
.02
9.7 (4.0) 0.72
.48
0.47
.64
2.28
.03 Note: Significant contrasts are
1.57 highlighted in bold
.13
Indirect
Speech
Table II. Numbers of no-regressions go-past fixations launched in Experiment 10 (with standard
deviations in brackets). Also shown are results from 2-tailed paired sample t-tests examining the simple
effects of Context and Quoting Style.
Direct
Speech
Fast-speaking Context
Slow-speaking Context
t(47)
By
Subjects
P
Context
Contrast
t(23)
By
Items
P
9.1 (3.3)
9.7 (3.7)
-2.41
.02
-1.62
.12
Quoting Style Contrast
By Subjects
By Items
t(47)
p
t(23)
p
9.7 (3.9) -2.84
.007 -2.14
.04
9.3 (3.7) 0.98
.33
0.12
.91
2.57
.01 Note: Significant contrasts are
1.21 highlighted in bold
.24
Indirect
Speech
As we did in the main analyses in the paper, we ran the same analyses again, but with
items 1, 23, and 24 removed. The Context  Quoting Style interaction remained reliable in
numbers of go-past fixations (F1(1, 47) = 7.61, p < .01, F2(1, 20) = 4.75, p = .04) and, at
least marginally, in numbers of no-regressions go-past fixations (F1(1, 47) = 4.82, p = .03,
F2(1, 20) = 3.25, p < .09) – importantly, just as with the reading time measures in the paper,
Page | 186
the Context simple effect in the indirect speech condition (cf. Tables I and II) disappeared
when those three items were removed (Tables III and IV).
Table III. Numbers of go-past fixations launched in Experiment 10 (with standard deviations in brackets)
after removing items 1, 23, and 24 from analysis. Also shown are results from 2-tailed paired sample ttests examining the simple effects of Context and Quoting Style.
Direct
Speech
Fast-speaking Context
Slow-speaking Context
t(47)
By
Subjects
P
Context
Contrast
t(20)
By
Items
p
9.8 (3.7)
10.4 (3.9)
-2.82
.007
-1.75
.10
Quoting Style Contrast
By Subjects
By Items
t(47)
p
t(20)
p
10.3 (4.2) -1.62
.11 -2.00
.06
10.0 (4.1) 0.97
.34
1.01
.32
0.42
.68 Note: Significant contrasts are
0.41 highlighted in bold
.69
Indirect
Speech
Table IV. Numbers of no-regressions go-past fixations launched in Experiment 10 (with standard
deviations in brackets) after removing items 1, 23, and 24 from analysis. Also shown are results from 2tailed paired sample t-tests examining the simple effects of Context and Quoting Style.
Direct
Speech
Fast-speaking Context
Slow-speaking Context
t(47)
By
Subjects
P
Context
Contrast
t(20)
By
Items
p
(ii)
9.5 (3.4)
10.0 (3.7)
-2.36
.02
-1.63
.12
Quoting Style Contrast
By Subjects
By Items
t(47)
p
t(20)
p
9.8 (4.0) -1.97
.06 -1.77
.09
9.5 (3.9) 1.86
.07
0.85
.41
0.84
.41 Note: Significant contrasts are
0.12 highlighted in bold
.91
Indirect
Speech
Average durations of fixations
The corresponding analyses on average durations of go-past (respectively no-regressions
go-past) fixations revealed no reliable effects whatsoever, neither with the full set of items
(all ps ≥ .15; maximum between-condition difference: 4 ms) nor when items 1, 23, and 24
were removed (all ps > .20; maximum between-condition difference: 5 ms).
Discussion
Taken together, these supplementary analyses suggest that the reading time effects reported
in the paper are mainly due to modulations in the numbers of go-past fixations launched on
the critical regions – average durations of those fixations were not affected much by the
experimental conditions. It will be an interesting and important challenge for future
research to pin down exactly what kind of eye-movement control mechanisms are involved
in those pragmatically induced reading-rate modulations (e.g., whether they also involve
changes in low-level word processing). However, such future research will almost
certainly have to consider more measures and experimental conditions than we were able
to include at present.
Page | 187
Appendix 7 Reading materials for Experiment 11
[D] = Direct speech; [I] = Indirect speech; [FF] = face focused; [FU] = face unfocused
The critical sentences (for acoustic analyses in oral reading) are in bold and italic text
Happy stories
1.
Escaping from the hustle and bustle of the crowded city, couple Leona and Jason were enjoying an afternoon walk
in the park. Leona revelled in this peaceful place as she saw baby ducks following their mother in the pond.
[FF – D] She leaned on Jason with a satisfied smile: „Aww, this is so relaxing after a busy week. It's almost
magical!‟
[FF- I] She leaned on Jason with a satisfied smile, saying that it was so relaxing after a busy week and that it was
almost magical.
[FU – D] Leaning over, she gave Jason a big hug: „Aww, this is so relaxing after a busy week. It's almost magical!‟
[FU – I] Leaning over, she gave Jason a big hug, saying that it was so relaxing after a busy week and that it was
almost magical.
2.
Derek‟s birthday was just a couple of days away and his girlfriend Ruby had arranged something special. A few
days before, Ruby decided to surprise him with the plan that she kept hidden from him for so long.
[FF- D] Grinning from ear to ear, she revealed: „Look, darling, your birthday is coming up soon, so…I booked us
tickets to London to visit the opera!‟
[FF – I] Grinning from ear to ear, she revealed that since his birthday was coming up soon, she booked them
tickets to London to visit the opera.
[FU – D] Stroking his cheek devotedly, she revealed: „Look, darling, your birthday is coming up soon, so…I
booked us tickets to London to visit the opera!‟
[FU – I] Stroking his cheek devotedly, she revealed that since his birthday was coming up soon, she booked them
tickets to London to visit the opera.
3.
In her dinner party, Vicky was a bit nervous because she had prepared some exotic recipes that she had never
attempted before. Unexpectedly, all dishes went extremely popular and some guests even craved more after they
finished.
[FF – D] Vicky smiled jubilantly: „Wow, I didn’t expect that. I am so glad that you all enjoyed it!‟
[FF – I] Vicky smiled jubilantly, saying that she did not expect that and she was so glad that they all enjoyed it.
[FU – D] Vicky felt rather relieved: „Wow, I didn’t expect that. I am so glad that you all enjoyed it!‟
[FU – I] Vicky felt rather relieved, saying that she did not expect that and she was so glad that they all enjoyed it.
4.
In the small town of Oban lies a charming little pub. One evening, people were having a peaceful and relaxing night
as usual when the famous writer Aileen burst in with exhilaration.
[FF – D] She cheered with a big smile on her face: „Guess what – my new novel has been accepted for publication!
I'll buy everyone around.‟
[FF – I] She cheered with a big smile on her face, announcing that her new novel had been accepted for
publication and she would buy everyone a round.
[FU – D] Jumping up and down, she cheered: „Guess what – my new novel has been accepted for publication! I'll
buy everyone around.‟
[FU – I] Jumping up and down, she cheered, announcing that her new novel had been accepted for publication and
she would buy everyone a round.
5.
David was one of the best students in his school but he might not be able to go to university because of his poor
family. He made a few applications, hoping that he might get some financial support or a scholarship and today he
received a letter from Cambridge.
[FF – D] Opening the envelope, he grinned widely: „I got a full scholarship from University of Cambridge?! My
parents will be very proud.‟
[FF – I] Opening the envelope, he grinned widely, saying that he got a full scholarship from University of
Cambridge and his parents would be very proud.
[FU – D] He thrilled to pieces as he opened the envelope, shouting: „I got a full scholarship from University of
Cambridge?! My parents will be very proud!‟
[FU – I] He thrilled to pieces as he opened the envelope, shouting that he got a full scholarship from University of
Cambridge and his parents would be very proud.
Page | 188
6.
The girls felt ravenous at 3am after hours of dancing and were desperately searching for a take away. After a few
blocks, they finally gorged themselves in a chip shop and were now laughing at each other‟s faces as they ate.
[FF – D] With a hiccup, Hannah giggled: „Oh, I love chips! They are just heavenly!‟
[FF – I] With a hiccup, Hannah giggled, saying that she loved chips and they were just heavenly.
[FU – D] With a hiccup, Hannah stroked her tummy contentedly: „Oh, I love chips! They are just heavenly!‟
[FU – I] With a hiccup, Hannah stroked her tummy contentedly, saying that she loved chips and they were just
heavenly.
7.
It was the last day of John Lewis‟s summer sale when two girls Helen and Abbey were browsing around. Helen was
very fond of a black skirt and tried it on straightaway when she found it was heavily discounted.
[FF – D] She was grinning widely: „Hey! How do I look? It is exactly what I always wanted, and now it is even
more than 50% off!‟
[FF – I] Grinning widely, she asked how she looked and said that it was exactly what she always wanted, and that
now it was even more than 50% off.
[FU – D] Jumping up and down, she showed it to Abbey: „Hey! How do I look? It is exactly what I always wanted,
and now it is even more than 50% off!‟
[FU – I] Jumping up and down, she showed it to Abbey, asking how she looked and said that it was exactly what
she always wanted, and that now it was even more than 50% off.
8.
Business woman Stella had a rather sunny and cheerful day because she was just promoted to the department
manager. When she got home, she was surprised that her husband had already prepared a candle-light dinner to
congratulate her.
[FF – D] She hugged him with a contented smile: „Thank you, my darling. This is very sweet of you!‟
[FF – I] She hugged him with a contented smile, thank him and said that it was very sweet of him.
[FU – D] Contentedly, she gave him a big hug: „Thank you, my darling. This is very sweet of you!‟
[FU – I] Contentedly, she gave him a big hug to thank him, saying that it was very sweet of him.
9.
After 2 weeks of international conferences, Lizzi finally arrived home and her puppy dashed into her arms, licking
her face ecstatically. Lizzi‟s exhaustion was instantly wiped away by such a heart-warming welcome from her little
dog.
[FF – D] She giggled: „Aww darling. I missed you too! I dreamt of you every single night in the past two weeks.‟
[FF – I] She giggled, saying that she missed her too and she dreamt about her every single night in the past two
weeks.
[FU – D] She stroked him devotedly: „Aww darling. I missed you too! I dreamt of you every single night in the
past two weeks.‟
[FU – I] She stroked him devotedly, saying that she missed her too and she dreamt about her every single night in
the past two weeks.
10. It was 25th Dec and school girl Melissa got up at 7am because she was eager to find out what she had received from
Santa Claus. A feeling of pure joy came over her as she opened the big pink box – a Barbie doll called 'Emma'!
[FF – D] She ran upstairs with a big smile on her face: „Mummy, Santa Claus has sent me a sister and her name is
Emma!‟
[FF – I] She ran upstairs with a big smile on her face, telling her mum that Santa Claus had sent her a sister and
her name was Emma.
[FU – D] She ran upstairs, shouting joyfully: „Mummy, Santa Claus has sent me a sister and her name is Emma!‟
[FU – I] She ran upstairs, shouting joyfully to her mum that Santa Claus had sent her a sister and her name was
Emma.
11. PhD student Ella had just finished her viva and was now wandering nervously outside the exam room, waiting for
the final decision. Her friends and colleagues had gathered to support her until the examiners came out with a round
of applause.
[FF – D] Over the moon, Ella was grinning from ear to ear: „Oh, I would not have gone through this without you,
my friends! Now, it is time for us to celebrate!‟
[FF – I] Over the moon, Ella was grinning from ear to ear, saying that she would not have gone that far without
them and now it was time for them to celebrate.
[FU – D] Thrilled to pieces, Ella hugged her friends joyously: „Oh, I would not have gone through this without you,
my friends! Now, it is time for us to celebrate!‟
[FU – I] Thrilled to pieces, Ella hugged her friends joyously, saying that she would not have gone that far without
them and now it was time for them to celebrate.
Page | 189
12. It was approaching the end of the term; first-year student Katie was withdrawing her essay mark in the lab. Although
she had worked very hard, she was slightly worried because many students had received a C.
[FF – D] She smiled jubilantly as she saw an A-: „Oh thank god. I was very worried. I am so glad that my efforts
paid off!‟
[FF – I] She smiled jubilantly as she saw an A-, saying that she had been very worried and she was so glad that her
efforts had paid off.
[FU – D] She felt relieved as she saw an A-: „Oh thank god. I was very worried. I am so glad that my efforts paid
off!‟
[FU – I] She felt relieved as she saw an A-, saying that she had been very worried and she was so glad that her
efforts had paid off.
Surprised Stories (positive)
13. Alex was a world-renowned art collector who was specialized in Italian renaissance paintings. He bought a secondhand painting box for his granddaughter and was shocked when he discovered the inscription: „1472, Leonardo da
Vinci‟ at the inner edge of the box.
[FF – D] His eyes became wide open: „Oh my goodness, I can’t believe this is da Vinci’s painting box!‟
[FF – I] Eyes wide open, he said that he couldn't believe it was da Vinci’s painting box.
[FU – D] He was completely gobsmacked: „Oh my goodness, I can’t believe this is da Vinci’s painting box!‟
[FU – I] He was completely gobsmacked, saying that he couldn't believe it was da Vinci’s painting box.
14. Conservation students Sarah and Brian drifted along the coast in their sea kayak, spotting local birds and wildlife.
As they rounded a cliff, they came across a breath-taking Sanford‟s Sea-eagle which was only 5 meters away from
them.
[FF – D] Brian's mouth fell open in surprise: „Bloody hell! This is a Sanford’s Sea-eagle! It has never been found
in Britain before!‟
[FF – I] Brian's mouth fell open in surprise and he said that it was a Sanford’s Sea-eagle and it had never been
found in Britain before.
[FU – D] Holding his breath, Brian was shocked: „Bloody hell! This is a Sanford’s Sea-eagle! It has never been
found in Britain before!‟
[FU – I] Holding his breath, Brian was shocked, saying that it was a Sanford’s Sea-eagle and it had never been
found in Britain before.
15. In Beijing, tourists Alison and Nick were going to a big silk market because they had never tried silk clothes before.
When they entered the market, they found themselves overwhelmed by a wide range of fine fabric on offer.
[FF – D] Alison touched one and her eyes were instantly widened: „Nick! The silk feels so incredibly smooth!‟
[FF – I] As Alison touched one, her eyes were instantly widened and she told Nick that the silk felt so incredibly
smooth.
[FU – D] Alison's voice was filled with surprise: „Nick! The silk feels so incredibly smooth!‟
[FU – I] Alison's voice was filled with surprise as she told Nick that the silk felt so incredibly smooth.
16. Dave had been a lucky postman because he had won a fortune in the lottery and soon purchased an expensive Ferrari
328GTS Targa. Today in the car park, his colleague Ralph was utterly shocked by Dave‟s new possession.
[FF – D] His mouth fell open in surprise: „Wow! Is that your new car!? It looks really stylish!‟
[FF – I] His mouth fell open in surprise as he asked Dave whether that was his new car, adding that it looked
really stylish.
[FU – D] His voice filled with envy: „Wow! Is that your new car!? It looks really stylish!‟
[FU – I] His voice filled with envy, asking Dave whether that was his new car, adding that it looked really stylish.
17. Peter was upset because his dad - his best mate - did not give him any presents for his 18th birthday today. He went
to the garage to fetch his soccer ball and was rooted to the spot as he saw a brand new Toyota inscribed with „Peter's
18th, love dad‟.
[FF – D] His eyes became wide open: „Good lord! I’m not dreaming, am I? It looks absolutely amazing!‟
[FF – I] Eyes wide open, he asked if he was dreaming, adding that it looked absolutely amazing.
[FU – D] He was completely gobsmacked: „Good lord! I’m not dreaming, am I? It looks absolutely amazing!‟
[FU – I] He was completely gobsmacked, asking if he was dreaming, adding that it looked absolutely amazing.
Page | 190
18. In MRC biomedical laboratory, Dr. Parkinson was browsing the microscopic photos taken in their recent
experiments. As he saw the effects of peptide-24 on amyloid β proteins on the infected brain tissues, he was utterly
astonished.
[FF – D] His eyes were widened: „A… A significant shrinkage! This could be a potential cure for Alzheimer’s
disease!‟
[FF – I] Eyes widened, he said that there was a significant shrinkage and that it could be a potential cure for
Alzheimer’s disease.
[FU – D] He breathed rapidly: „A… A significant shrinkage! This could be a potential cure for Alzheimer’s
disease!‟
[FU – I] He breathed rapidly, saying that there was a significant shrinkage and that it could be a potential cure for
Alzheimer’s disease.
Surprise stories (negative)
19. At the beach, William and Phil were enjoying the sunshine and a fresh sea breeze. Suddenly, William noticed a
piece of wood floating on the water with something lying on it that looked like an unconscious child.
[FF – D] Eyes wide open, he gasped: „Phil, quick! We need to call for help because there is a child in trouble!‟
[FF – I] Eyes wide open, he gasped and told Phil that they needed to call for help because there was a child in
trouble.
[FU – D] He was numb with shock: „Phil, quick! We need to call for help because there is a child in trouble!‟
[FU – I] He was numb with shock, telling Phil that they needed to call for help because there was a child in
trouble.
20. At a peaceful house in Dumbarton Mrs. Harris was preparing lunch for her children. The phone rang and Mrs. Harris
was stunned by what she heard on the phone.
[FF – D] Her face looked rather frightened and she said in horror: „Oh no! My husband was involved in a collision
with a heavy goods vehicle and is now in hospital!‟
[FF – I] Her face looked rather frightened and she said in horror that her husband was involved in a collision with
a heavy goods vehicle and was now in hospital.
[FU – D] She was dazed at the scene, muttering: „Oh no! My husband was involved in a collision with a heavy
goods vehicle and is now in hospital!‟
[FU – I] She was dazed at the scene, muttering that her husband was involved in a collision with a heavy goods
vehicle and was now in hospital.
21. In the library, university student Suzanne was having a snack break from busy exam revision. After two packs of
Milk Chocolate-Covered Oreos, she instantly gasped when she read the back of the package.
[FF – D] Her eyes nearly came out of her sockets: „Oh no!! What have I done! I just had a 2000-calorie snack!‟
[FF – I] Her eyes nearly came out of her sockets while she asked herself what she had done, adding that she just
had a 2000-calorie snack.
[FU – D] She was grabbing her hair: „Oh no!! What have I done! I just had a 2000-calorie snack!‟
[FU – I] She was grabbing her hair while she asked herself what she had done, adding that she just had a 2000calorie snack.
22. There had been a couple of complaints about the food in the 1-Michelin star restaurant Alain Ducasse this evening.
The restaurant manager came into the kitchen for a look and was instantly astounded by its unacceptable state.
[FF – D] His face looked rather frightened: „This cannot be true! The fish is burnt, the vegetables overcooked, and
the soup is actually stone cold...‟
[FF – I] His face looked rather frightened as he said that it could not be true that the fish was burnt, the vegetables
overcooked, and the soup was actually stone cold.
[FU – D] He was grabbing his hair: „This cannot be true! The fish is burnt, the vegetables overcooked, and the
soup is actually stone cold...‟
[FU – I] He was grabbing his hair, saying that it could not be true that the fish was burnt, the vegetables
overcooked, and the soup was actually stone cold.
23. It was one week before Christmas and Elaine was still struggling to finish all her work before her holiday. Her
glimmer of hope for a relaxing holiday was now mercilessly destroyed by yet another huge pile of files on her desk.
[FF – D] Her eyes nearly came out of her sockets: „I don’t believe this. There seems to be no end to it in sight...‟
[FF – I] Her eyes nearly came out of her sockets while she said she couldn't believe it and that there seemed to be
no end to it in sight.
[FU – D] Looking at the pile, Elaine was numb with shock: „I don’t believe this. There seems to be no end to it in
sight...‟
[FU – I] Looking at the pile, Elaine was numb with shock, saying that she couldn't believe it and that there seemed
to be no end to it in sight.
Page | 191
24. English fan John's blissful happiness from Steven Gerrard‟s leading goal against USA did not last long. Shockingly,
the English goalkeeper Robert Green inexplicably allowed USA's soft, skidding shot to escape his grasp and roll into
the net.
[FF – D] Eyes wide open, he gasped: „I don’t believe this. Green, you must have butter fingers!‟
[FF – I] Eyes wide open, he gasped and said that he did not believe what had happened and that Green must have
butter fingers.
[FU – D] John was dazed at the scene, muttering: „I don’t believe this. Green, you must have butter fingers!‟
[FU – I] John was dazed at the scene, muttering that he did not believe what had happened and that Green must
have butter fingers.
Angry stories
25. First-year university student Britney found that the weather in Glasgow was one of the worst in Britain. After a
heavy snow in the afternoon, she fell on her elbow and found that her new, expensive jacket was dirty and scratched.
[FF – D] She complained with an angry face: „I really hate the winter! It’s always dark and the roads are too
slippery.‟
[FF – I] She complained with an angry face that she really hated the winter because it was always dark and the
roads were too slippery.
[FU – D] She complained in an angry voice: „I really hate the winter! It’s always dark and the roads are too
slippery.‟
[FU – I] She complained in an angry voice that she really hated the winter because it was always dark and the
roads were too slippery.
26. Universities in England were preparing to raise tuition fees following the government's cuts of 40%.Simon, among
thousands of students, was angrily protesting outside Parliament today. When he was interviewed, he appeared
particularly indignant.
[FF – D] He frowned disapprovingly: „You know what, the MPs have broken their promises, peddle lies about
fairness and we will expose their betrayal!‟
[FF – I] He frowned disapprovingly, saying that the MPs had broken their promises, peddled lies about fairness
and that the students would expose their betrayal.
[FU – D] He aired his grievances: „You know what, the MPs have broken their promises, peddle lies about
fairness and we will expose their betrayal!‟
[FU – I] He aired his grievances, saying that the MPs had broken their promises, peddled lies about fairness and
that the students would expose their betrayal.
27. The nursery was filled with lots of young children, screaming for attention from the busy childminder. One boy in
particular had been told off many times for behaving disgracefully all day and yet he was still throwing sponge balls
at the poor lady.
[FF – D] With an annoyed frown, she stressed to him: „Listen. You have to be really good and not make a mess for
the next few hours.‟
[FF – I] With an annoyed frown, she stressed to him that he had to be really good and not make a mess for the
next few hours.
[FU – D] Pointing at the boy's nose, she exploded: „Listen. You have to be really good and not make a mess for the
next few hours.‟
[FU – I] Pointing at the boy's nose, she exploded, saying that he had to be really good and not make a mess for the
next few hours.
28. A trendy night club had been recently opened near Marvin‟s flat and it proved instantly popular. Every night,
Marvin was terribly disturbed at sleep by the thudding noise and this morning he was venting his anger to his
neighbour.
[FF – D] His face filled with resentment: „This club is ridiculous! Every time I have to work early hours, they are
having a bash!‟
[FF – I] His face was filled with resentment while he said that the club was ridiculous because every time he had to
work early hours, they were having a bash.
[FU – D] He was blustering: „This club is ridiculous! Every time I have to work early hours, they are having a
bash!‟
[FU – I] He was blustering that the club was ridiculous because every time he had to work early hours, they were
having a bash.
Page | 192
29. Jenny and Irvin were taking a holiday trip to Barcelona and they were very glad that they had booked a very cheap
flight. However, it turned out that they had to pay extra money for virtually every service on plane, which even
included £3 for the toilet.
[FF – D] Her face filled with rage: „What!? I have to pay for a pee? How ridiculous! This flight is an absolute ripoff!‟
[FF – I] Her face filled with rage while she said that it was ridiculous for her to pay for a pee, concluding that the
flight was an absolute rip-off.
[FU – D] She became increasingly cross, shouting: „What!? I have to pay for a pee? How ridiculous! This flight is
an absolute rip-off!‟
[FU – I] She became increasingly cross, shouting that it was ridiculous for her to pay for a pee, concluding that the
flight was an absolute rip-off.
30. James was annoyed when an important mail he had asked to be sent urgently was returned to the company in the
post after two weeks. He was in a foul mood and stormed through to his assistant‟s office, not bothering to knock on
the door.
[FF – D] Banging the package on the desk, James gnashed his teeth: „If you don’t want to be fired, make sure it
gets to the right person this time!‟
[FF – I] Banging the package on the desk, James gnashed his teeth, saying that if the assistant did not want to be
fired, he should make sure it got to the right person this time.
[FU – D] Banging the package on the desk, James shouted furiously: „If you don’t want to be fired, make sure it
gets to the right person this time!‟
[FU – I] Banging the package on the desk, James shouted furiously that if the assistant did not want to be fired, he
should make sure it got to the right person this time.
31. Millionaire Joseph was addicted to betting on horses and he had always been dreaming of a big win. Tipped by a socalled „insider‟, he recently placed an enormous bet, but shockingly, the horse had lost.
[FF – D] His face was filled with resentment while he was on the phone: „Where did your bloody information come
from!? That was a huge amount of money – almost one million pounds!‟
[FF – I] His face was filled with resentment while he was on the phone, asking where his bloody information came
from, adding that it was a huge amount of money – almost one million pounds.
[FU – D] Angry with his informant, Joseph shouted furiously on the phone: „Where did your bloody information
come from!? That was a huge amount of money – almost one million pounds!‟
[FU – I] Angry with his informant, Joseph shouted furiously on the phone asking where his bloody information
came from, adding that it was a huge amount of money – almost one million pounds.
32. Leanne‟s son had promised her that he would tidy his room when he got home from school. She was furious to find
him watching a film, his speakers blaring and his room still a mess.
[FF – D] She lunged to turn the TV off and gnashed her teeth: „That’s it! You’re going to be grounded until this
room is sparkling.‟
[FF – I] She lunged to turn the TV off and gnashed her teeth, saying that was it and he was going to be grounded
until that room was sparkling.
[FU – D] She lunged to turn the TV off and exploded: „That’s it! You’re going to be grounded until this room is
sparkling.‟
[FU – I] She lunged to turn the TV off and exploded, saying that was it and he was going to be grounded until that
room was sparkling.
33. It was Sunday afternoon when Heather finally came back home after visiting her parents. She was outraged to find
her boyfriend Ben sleeping like a pig in his messy bedroom where she couldn‟t even find a space to stand.
[FF – D] She dragged Ben off the bed and roared with an angry face: „I can’t believe how messy you are! The first
thing I found on the floor were your dirty pyjamas!‟
[FF – I] She dragged Ben off the bed and roared with an angry face, saying that she could not believe how messy he
was and that the first thing she found on the floor were his dirty pyjamas.
[FU – D] She dragged Ben off the bed and roared in his face: „I can’t believe how messy you are! The first thing I
found on the floor were your dirty pyjamas!‟
[FU – I] She dragged Ben off the bed and roared in his face, saying that she could not believe how messy he was
and that the first thing she found on the floor were his dirty pyjamas.
Page | 193
34. During a coffee break, university teachers Robert and Isabel met in the common room. Robert was irritated with the
mess in the kitchen and according to a post-it, his mug had even been „borrowed‟ by someone else.
[FF – D] His face was filled with rage: „How brazen are these people! They didn’t want to wash their dirty mugs
so they just 'borrowed' mine!‟
[FF – I] His face was filled with rage as he said that these people were very brazen because when they didn't want
to wash their dirty mugs, they just 'borrowed' his.
[FU – D] He aired his grievances to Isabel: „How brazen are these people! They didn’t want to wash their dirty
mugs so they just 'borrowed' mine!‟
[FU – I] He aired his grievances to Isabel saying that these people were very brazen because when they didn't want
to wash their dirty mugs, they just 'borrowed' his.
35. Shauna was a young, talented actress and singer but she had been unemployed for months. She was now desperate
for a part in a new musical project but her boyfriend was ill and blamed her for not looking after him.
[FF – D] With an annoyed frown, Shauna stressed to him: „Don‟t be ridiculous! I am doing this because without
this job I can’t even pay for the electricity bill!‟
[FF – I] With an annoyed frown, Shauna stressed to him that he was ridiculous and she was doing that because
without that job she couldn’t even pay for the electricity bill.
[FU – D] Shauna was angry with his attitude, blustering: „Don‟t be ridiculous! I am doing this because without this
job I can’t even pay for the electricity bill!‟
[FU – I] Shauna was angry with his attitude, blustering that he was ridiculous and she was doing that because
without that job she couldn’t even pay for the electricity bill.
36. One morning, six-year-old Sara was taking a lot of time to get ready for school and the school bus was due to arrive
soon. Although her dad had been reminding her for half hour, she was still in her pyjamas.
[FF – D] Her dad frowned disapprovingly: „Sara! If you are not ready within five minutes, I’ll have to drive you
again!‟
[FF – I] Her dad frowned disapprovingly, saying that if she were not ready within five minutes, he would have to
drive her again.
[FU – D] Her dad became increasingly cross, shouting: „Sara! If you are not ready within five minutes, I’ll have to
drive you again!‟
[FU – I] Her dad became increasingly cross, shouting that if she were not ready within five minutes, he would have
to drive her again.
Page | 194
References
Abramson, M. (2007). The written voice: Implicit memory effects of voice characteristics
following silent reading and auditory presentation. Perceptual and Motor Skills, 105(3),
1171-1186.
Abramson, M., & Goldinger, S. D. (1997). What the reader's eye tells the mind's ear: Silent reading
activates inner speech. Perception & Psychophysics, 59(7), 1059-1068.
Alexander, J. D., & Nygaard, L. C. (2008). Reading voices and hearing text: Talker-specific
auditory imagery in reading. Journal of Experimental Psychology-Human Perception and
Performance, 34(2), 446-459.
Anderson, J. R. (1974). Verbatim and propositional representation of sentences in immediate and
long-term memory. Journal of Verbal Learning and Verbal Behavior, 13(2), 149-162.
Ashby, J., & Clifton, C. (2005). The prosodic property of lexical stress affects eye movements
during silent reading. Cognition, 96(3), B89-B100.
Ashby, J., & Rayner, K. (2004). Representing syllable information during silent reading: Evidence
from eye movements. Language and Cognitive Processes, 19(3), 391-426.
Banse, R., & Scherer, K. R. (1996). Acoustic profiles in vocal emotion expression. Journal of
Personality and Social Psychology, 70(3), 614-636.
Banziger, T., & Scherer, K. R. (2005). The role of intonation in emotional expressions. Speech
Communication, 46(3-4), 252-267.
Barsalou, L. W. (1999). Perceptual symbol systems. Behavioral and Brain Sciences, 22(4), 577609.
Barsalou, L. W. (2008). Grounded cognition. Annual Review of Psychology, 59, 617-645.
Barsalou, L. W. (2009). Simulation, situated conceptualization, and prediction. Philosophical
Transactions of the Royal Society B-Biological Sciences, 364(1521), 1281-1289.
Barsalou, L. W. (2010). Grounded Cognition: Past, Present, and Future. Topics in Cognitive
Science, 2(4), 716-724.
Beattie, G., Cutler, A., & Pearson, M. (1982). „Why is Mrs Thatcher interrupted so often?‟. Nature,
300, 744-747.
Belin, P., Zatorre, R. J., Lafaille, P., Ahad, P., & Pike, B. (2000). Voice-selective areas in human
auditory cortex. Nature, 403(6767), 309-312.
Blyth, C., Recktenwald, S., & Wang, J. (1990). I'm like, „Say what?!‟: A new quotative in
American oral narrative. American Speech, 65(3), 215-227.
Boersma, P., & Weenink, D. (2010). Praat: doing phonetics by computer. Version 5.1.43, retrieved
4 August 2010 from http://www.praat.org/.
Bohan, J., Sanford, A. J., Cochrane, S., & Sanford, A. J. S. (2008). Direct and indirect speech
modulates depth of processing. Paper presented at the 14th Annual Conference on
Architectures and Mechanisms for Language Processing (AMLaP), Cambridge, UK.
Bolinger, D. (1964). Intonation as a universal. Paper presented at the Proceedings of the 5th
Congress of Phonetics.
Borreggine, K. L., & Kaschak, M. P. (2006). The action-sentence compatibility effect: It's all in the
timing. Cognitive Science, 30(6), 1097-1112.
Brown, C. A., & Bacon, S. P. (2010). Fundamental frequency and speech intelligibility in
background noise. Hearing Research, 266(1-2), 52-59.
Bubic, A., von Cramon, D. Y., Jacobsen, T., Schroger, E., & Schubotz, R. I. (2009). Violation of
Expectation: Neural Correlates Reflect Bases of Prediction. Journal of Cognitive
Neuroscience, 21(1), 155-168.
Bulut, M., & Narayanan, S. (2008). On the robustness of overall F0-only modifications to the
perception of emotions in speech. Journal of the Acoustical Society of America, 123(6),
4547-4558.
Bunzeck, N., Wuestenberg, T., Lutz, K., Heinze, H. J., & Jancke, L. (2005). Scanning silence:
Mental imagery of complex sounds. Neuroimage, 26(4), 1119-1127.
Page | 195
Busso, C., Lee, S., & Narayanan, S. (2009). Analysis of Emotionally Salient Aspects of
Fundamental Frequency for Emotion Detection. Ieee Transactions on Audio Speech and
Language Processing, 17(4), 582-596.
Calvert, G. A., Bullmore, E. T., Brammer, M. J., Campbell, R., Williams, S. C. R., McGuire, P. K.,
& David, A. S. (1997). Activation of auditory cortex during silent lipreading. Science,
276(5312), 593-596.
Chafe, W. L. (Ed.). (1982). Integration and involvement in speaking, writing, and oral literature:
Norwood, NJ: Ablex.
Chomsky, N. (1957). Syntactic Structures: The Hague: Mouton.
Clark, H. H., & Gerrig, R. J. (1990). Quotations as Demonstrations. Language, 66(4), 764-805.
Cohen, A., Collier, R., & Thart, J. (1982). Declination: Construct or intrinsic feature of speech
pitch? [Article]. Phonetica, 39(4-5), 254-273.
Cohen, A., & t' Hart, J. (1967). On the anatomy of intonation. Lingua, 19, 177-192.
Connell, L. (2007). Representing object colour in language comprehension. Cognition, 102(3),
476-485.
Coulmas, F. (Ed.). (1986). Direct and Indirect Speech. Berlin; New York; Amsterdam: Mouton de
Gruyter.
Cox, R. M., Alexander, G. C., Taylor, I. M., & Gray, G. A. (1997). The contour test of loudness
perception. Ear and Hearing, 18(5), 388-400.
Cukor-Avila, P. (2002). She say, she go, she be like: Verbs of quotation over time in African
American Vernacular English. American Speech, 77(1), 3-31.
Daniel, C. (2009). Embodiment of abstract concepts: Good and bad in right- and left-handers.
Journal of Experimental Psychology: General, 138(3), 351-367.
Danly, M., Cooper, W. E., & Shapiro, B. (1983). Fundamental-frequency, language processing, and
linguistic structure in Wernicke's aphasia. Brain and Language, 19(1), 1-24.
Davis, M. H., & Johnsrude, I. S. (2007). Hearing speech sounds: Top-down influences on the
interface between audition and speech perception. Hearing Research, 229(1-2), 132-147.
Dodell-Feder, D., Koster-Hale, J., Bedny, M., & Saxe, R. (2011). fMRI item analysis in a theory of
mind task. NeuroImage, 55(2), 705-712.
Dzhelyova, M. P., Ellison, A., & Atkinson, A. P. (2011). Event-related repetitive TMS reveals
distinct, critical roles for right OFA and bilateral posterior STS in judging the sex and
trustworthiness of faces. Journal of Cognitive Neuroscience, 23(10), 2782-2796.
Enrici, I., Adenzato, M., Cappa, S., Bara, B. G., & Tettamanti, M. (2011). Intention processing in
communication: A common brain network for language and gestures. Journal of Cognitive
Neuroscience, 23(9), 2415-2431.
Ferrara, K., & Bell, B. (1995). Sociolinguistic variation and discourse function of constructed
dialogue introducers - the case of be + like. American Speech, 70(3), 265-290.
Fincher-Kiefer, R. (2001). Perceptual components of situation models. Memory & Cognition, 29(2),
336-343.
Fischer, M. H., & Zwaan, R. A. (2008). Embodied language: A review of the role of the motor
system in language comprehension. Quarterly Journal of Experimental Psychology, 61(6),
825-850.
Fodor, J. A. (1975). The language of thought. Cambridge, MA: Harvard University Press.
Glenberg, A. M., & Kaschak, M. P. (2002). Grounding language in action. Psychonomic Bulletin &
Review, 9(3), 558-565.
Glenberg, A. M., Sato, M., Cattaneo, L., Riggio, L., Palumbo, D., & Buccino, G. (2008).
Processing abstract language modulates motor system activity. Quarterly Journal of
Experimental Psychology, 61(6), 905-919.
Goodell, E. W., & Sachs, J. (1992). Direct and indirect speech in English-speaking children's retold
narratives. Discourse Processes, 15(4), 395-422.
Grandjean, D., Sander, D., Pourtois, G., Schwartz, S., Seghier, M. L., Scherer, K. R., &
Vuilleumier, P. (2005). The voices of wrath: Brain responses to angry prosody in
meaningless speech. Nature Neuroscience, 8(2), 145-146.
Page | 196
Hardin, J. W., & Hilbe, J. M. (2003). Generalized Estimating Equations. London: Chapman &
Hall/CRC.
Harnad, S. (1990). The symbol grounding problem. Physica D, 42(1-3), 335-346.
Havas, D. A., Glenberg, A. M., Gutowski, K. A., Lucarelli, M. J., & Davidson, R. J. (2010).
Cosmetic use of Botulinum Toxin-A affects processing of emotional language.
Psychological Science, 21(7), 895-900.
Hird, K., & Kirsner, K. (2002). The relationship between prosody and breathing in spontaneous
discourse. Brain and Language, 80(3), 536-555.
Hjelmquist, E., & Gidlund, Å. (1985). Free recall of conversations. Text - Interdisciplinary Journal
for the Study of Discourse, 5(3), 169 - 186.
Holmes, K. J., & Lourenco, S. F. (2011). Common spatial organization of number and emotional
expression: A mental magnitude line. Brain and Cognition, 77(2), 315-323.
Holt, E. (1996). Reporting on talk: The use of direct reported speech in conversation. Research on
Language and Social Interaction, 29(3), 219-245.
Hooker, C. I., Paller, K. A., Gitelman, D. R., Parrish, T. B., Mesulam, M. M., & Reber, P. J. (2003).
Brain networks for analyzing eye gaze. Cognitive Brain Research, 17(2), 406-418.
Huettel, S. A., Mack, P. B., & McCarthy, G. (2003). Brain systems associated with violations of
expectation. Society for Neuroscience Abstract Viewer and Itinerary Planner, 2003,
Abstract No. 343.347.
Jancke, L., & Shah, N. J. (2004). 'Hearing' syllables by 'seeing' visual stimuli. European Journal of
Neuroscience, 19(9), 2603-2608.
Jarvella, R. J. (1971). Syntactic processing of connected speech. Journal of Verbal Learning and
Verbal Behavior, 10(4), 409-416.
Johnsrude, I. S., Penhune, V. B., & Zatorre, R. J. (2000). Functional specificity in the right human
auditory cortex for perceiving pitch direction. Brain, 123, 155-163.
Kanwisher, N., McDermott, J., & Chun, M. M. (1997). The fusiform face area: A module in human
extrastriate cortex specialized for face perception. Journal of Neuroscience, 17(11), 43024311.
Kaschak, M. P., & Borreggine, K. L. (2008). Temporal dynamics of the action-sentence
compatibility effect. Quarterly Journal of Experimental Psychology, 61(6), 883-895.
Kaschak, M. P., & Glenberg, A. M. (2000). Constructing meaning: The role of affordances and
grammatical constructions in sentence comprehension. Journal of Memory and Language,
43(3), 508-529.
Kaschak, M. P., Madden, C. J., Therriault, D. J., Yaxley, R. H., Aveyard, M., Blanchard, A. A., &
Zwaan, R. A. (2005). Perception of motion affects language processing. Cognition, 94(3),
B79-B89.
Kaup, B., Ludtke, J., & Zwaan, R. A. (2006). Processing negated sentences with contradictory
predicates: Is a door that is not open mentally closed? Journal of Pragmatics, 38(7), 10331050.
Kellenbach, M. L., Brett, M., & Patterson, K. (2001). Large, colorful, or noisy? Attribute- and
modality-specific activations during retrieval of perceptual attribute knowledge. Cognitive
Affective and Behavioral Neuroscience, 1(3), 207-221.
Kent, R. D., & Read, C. (1992). The acoustic analysis of speech. San Diego: Singular Publishing
Group.
Kiefer, M., & Pulvermüller, F. (in press). Conceptual representations in mind and brain:
Theoretical developments, current evidence and future directions. Cortex.
Kiefer, M., Sim, E. J., Herrnberger, B., Grothe, J., & Hoenig, K. (2008). The sound of concepts:
four markers for a link between auditory and conceptual brain systems. Journal of
Neuroscience, 28(47), 12224-12230.
Kintsch, W. (1998). Comprehension: A paradigm for cognition. Cambridge, MA: Cambridge
University Press.
Kintsch, W., & Vandijk, T. A. (1978). Toward a model of text comprehension and production.
Psychological Review, 85(5), 363-394.
Page | 197
Klein, M. E., & Zatorre, R. J. (2011). A role for the right superior temporal sulcus in categorical
perception of musical chords. Neuropsychologia, 49(5), 878-887.
Kosslyn, S. M., & Matt, A. M. C. (1977). If you speak slowly, do people read your prose slowly person-perticular speech recoding during reading. Bulletin of the Psychonomic Society,
9(4), 250-252.
Kring, A. M., & Gordon, A. H. (1998). Sex differences in emotion: Expression, experience, and
physiology. Journal of Personality and Social Psychology, 74(3), 686-703.
Kurby, C. A., Magliano, J. P., & Rapp, D. N. (2009). Those voices in your head: Activation of
auditory images during reading. Cognition, 112(3), 457-461.
Labov, W. (1972). The transformation of experience in narrative syntax. Language in the inner city
(pp. 354-396): Philadelphia: University of Pennsylvania Press.
Leech, G. N., & Short, M. H. (1981). Style in fiction: A linguistic introduction to English fictional
prose. London: Longman.
Lehrer, A. (1989). Remembering and representing prose - quoted speech as a data source.
Discourse Processes, 12(1), 105-125.
Lynott, D., & Connell, L. (2009). Modality exclusivity norms for 423 object properties. Behavioral
Research Methods, 41(2), 558-564.
MacSweeney, M., Amaro, E., Calvert, G. A., Campbell, R., David, A. S., McGuire, P., & Brammer,
M. J. (2000). Silent speechreading in the absence of scanner noise: an event-related fMRI
study. [Article]. NeuroReport, 11(8), 1729-1733.
Matsumoto, D., Keltner, D., Shiota, M. N., O'Sullivan, M., & Frank, M. (Eds.). (2008). Facial
Expressions of Emotion. New York; London: the Guilford Press.
Menn, L., & Boyce, S. (1982). Fundamental frequency and discourse structure. Language and
Speech, 25(4), 341-383.
Mitchell, R. L. (2006). How does the brain mediate interpretation of incongruent auditory emotions?
The neural response to prosody in the presence of conflicting lexico-semantic cues.
European Journal of Neuroscience, 24(12), 3611-3618.
Mitchell, R. L. C., Elliott, R., Barry, M., Cruttenden, A., & Woodruff, P. W. R. (2003). The neural
response to emotional prosody, as revealed by functional magnetic resonance imaging.
Neuropsychologia, 41(10), 1410-1421.
Morrison, R., & Rayner, K. (1981). Saccade size in reading depends upon character spaces and not
visual angle. Attention, Perception, & Psychophysics, 30(4), 395-396.
Niedenthal, P. M., Mermillod, M., Maringer, M., & Hess, U. (2010). The Simulation of Smiles
(SIMS) model: Embodied simulation and the meaning of facial expression. Behav Brain
Sci, 33(6), 417-433; discussion 433-480.
Ogawa, S., & Lee, T. M. (1990). Magnetic-resonance-imaging of blood-vessels at high fields: in
vivo and in vitro measurements and image stimulation. Magnetic Resonance in Medicine,
16(1), 9-18.
Ogawa, S., Lee, T. M., Kay, A. R., & Tank, D. W. (1990). Brain magnetic-resonance-imaging with
contrast dependent on blood oxygenation. Proceedings of the National Academy of
Sciences of the United States of America, 87(24), 9868-9872.
Ogawa, S., Lee, T. M., Nayak, A. S., & Glynn, P. (1990). Oxygenation-sensitive contrast in
magnetic-resonance image of rodent brain at high magnetic-fields. Magnetic Resonance in
Medicine, 14(1), 68-78.
Pecher, D., Zeelenberg, R., & Barsalou, L. W. (2003). Verifying different-modality properties for
concepts produces switching costs. Psychological Science, 14(2), 119-124.
Peelen, M. V., Atkinson, A. P., & Vuilleumier, P. (2010). Supramodal Representations of
Perceived Emotions in the Human Brain. Journal of Neuroscience, 30(30), 10127-10134.
Pell, M. D. (2001). Influence of emotion and focus location of prosody in matched statements and
questions. Journal of the Acoustical Society of America, 109(4), 1668-1680.
Pitcher, D., Dilks, D. D., Saxe, R. R., Triantafyllou, C., & Kanwisher, N. (2011). Differential
selectivity for dynamic versus static information in face-selective cortical regions.
Neuroimage, 56(4), 2356-2363.
Page | 198
Pitt, M. A., & Crowder, R. G. (1992). The role of spectral and dynamic cues in imagery for musical
timbre. Journal of Experimental Psychology: Human Perception and Performance, 18(3),
728-738.
Pulvermuller, F. (2005). Brain mechanisms linking language and action. Nature Reviews
Neuroscience, 6(7), 576-582.
Pylyshyn, Z. W. (1984). Computation and cognition: Toward a foundation for cognitive science:
Cambridge, MA: MIT Press.
Richardson, D. C., & Spivey, M. J. (2000). Representation, space and Hollywood Squares: looking
at things that aren't there anymore. Cognition, 76(3), 269-295.
Richardson, D. C., Spivey, M. J., Barsalou, L. W., & McRae, K. (2003). Spatial representations
activated during real-time comprehension of verbs. Cognitive Science, 27(5), 767-780.
Robinson, M. D., & Johnson, J. T. (1997). Is it emotion or is it stress? Gender stereotypes and the
perception of subjective experience. Sex Roles, 36(3-4), 235-258.
Roy, C. S., & Sherrington, C. S. (1890). On the regulation of the blood-supply of the brain. The
Journal of physiology, 11(1-2), 85-158.117.
Russell, J. A., Bachorowski, J. A., & Fernandez-Dols, J. M. (2003). Facial and vocal expressions of
emotion. Annual Review of Psychology, 54, 329-349.
Sachs, J. S. (1967). Recognition memory for syntactic and semantic aspects of connected discourse.
Perception & Psychophysics, 2(9), 437-442.
Schiffrin, D. (1981). Tense variation in narrative. Language, 57(1), 45-62.
Scott, S. K., Blank, C. C., Rosen, S., & Wise, R. J. S. (2000). Identification of a pathway for
intelligible speech in the left temporal lobe. Brain, 123, 2400-2406.
Searle, J. R. (1980). Minds, Brains, and Programs. Behavioral and Brain Sciences, 3(3), 417-425.
Shergill, S. S., Bullmore, E. T., Brammer, M. J., Williams, S. C. R., Murray, R. M., & McGuire, P.
K. (2001). A functional study of auditory verbal imagery. Psychological Medicine, 31(2),
241-253.
Short, M., Semino, E., & Wynne, M. (2002). Revisiting the notion of faithfulness in discourse
presentation using a corpus approach. Language and Literature, 11(4), 325 - 355.
Sokolov, A. A., Erb, M., Gharabaghi, A., Grodd, W., Tatagiba, M. S., & Pavlova, M. A. (2011).
Biological motion processing: The left cerebellum communicates with the right superior
temporal sulcus. Neuroimage.
Spitsyna, G., Warren, J. E., Scott, S. K., Turkheimer, F. E., & Wise, R. J. S. (2006). Converging
language streams in the human temporal lobe. Journal of Neuroscience, 26(28), 7328-7336.
Stafford, L., Burggraf, C. S., & Sharkey, W. F. (1987). Conversational memory - the effects of time,
recall, mode, and memroy expectancies on remembrances of natural conversations. Human
Communication Research, 14(2), 203-229.
Stanfield, R. A., & Zwaan, R. A. (2001). The effect of implied orientation derived from verbal
context on picture recognition. Psychological Science, 12(2), 153-156.
Sturt, P., Sanford, A. J., Stewart, A., & Dawydiak, E. (2004). Linguistic focus and good-enough
representations: An application of the change-detection paradigm. Psychonomic Bulletin &
Review, 11(5), 882-888.
Tagliamonte, S., & D'Arcy, A. (2004). He's like, she's like: The quotative system in Canadian
youth. Journal of Sociolinguistics, 8(4), 493-514.
Tannen, D. (1982). Oral and Literate Strategies in Spoken and Written Narratives. Language, 58(1),
1-21.
Tannen, D. (1986). Introducing constructed dialogue in Greek and American conversational and
literary narrative. In F. Coulmas (Ed.), Direct and Indirect Speech. Berlin; New York;
Amsterdam: Mouton de Gruyter.
Tannen, D. (1989). „Oh talking voice that is so sweet‟: Constructing dialogue in conversation. In D.
Tannen (Ed.), Talking voices: Repetition, dialogue, and imagery in conversational
discourse. Cambridge: Cambridge University Press.
Taylor, L. J., & Zwaan, R. A. (2008). Motor resonance and linguistic focus. Quarterly Journal of
Experimental Psychology, 61(6), 896-904.
Page | 199
Thomson, A. J., & Martinet, A. V. (1986). A practical English grammar (4th ed.). Oxford: Oxford
University Press.
Tree, J. E. F., & Tomlinson, J. M. (2008). The rise of like in spontaneous quotations. Discourse
Processes, 45(1), 85-102.
Umeda, N. (1981). Influence of segmental factors on fundamental-frequency in fluent speech.
Journal of the Acoustical Society of America, 70(2), 350-355.
Vaissiere, J. (Ed.). (1983). Language-independent prosodic features. New York: Springer-Verlag.
Van Essen, D. C. (2002). Windows on the brain: the emerging role of atlases and databases in
neuroscience. Current Opinion in Neurobiology, 12(5), 574-579.
Van Essen, D. C., Drury, H. A., Dickson, J., Harwell, J., Hanlon, D., & Anderson, C. H. (2001). An
integrated software suite for surface-based analyses of cerebral cortex. Journal of the
American Medical Informatics Association, 8(5), 443-459.
Van Essen, D. C., Harwell, J., Hanlon, D., & Dickson, J. (Eds.). (2005). Surface-Based Atlases and
a Database of Cortical Structure and Function. New Jersey: John Wiley & Sons.
Wade, E., & Clark, H. H. (1993). Reproduction and demonstration in quotations. Journal of
Memory and Language, 32(6), 805-819.
Wiethoff, S., Wildgruber, D., Kreifelts, B., Becker, H., Herbert, C., Grodd, W., & Ethofer, T.
(2008). Cerebral processing of emotional prosody - influence of acoustic parameters and
arousal. Neuroimage, 39(2), 885-893.
Wildgruber, D., Ackermann, H., Kreifelts, B., & Ethofer, T. (2006). Cerebral processing of
linguistic and emotional prosody: fMRI studies. In S. Anders, G. Ende, M. Junghoffer, J.
Kissler & D. Wildgruber (Eds.), Understanding Emotions (Vol. 156, pp. 249-268).
Amsterdam: Elsevier Science Bv.
Wildgruber, D., Riecker, A., Hertrich, I., Erb, M., Grodd, W., Ethofer, T., & Ackermann, H. (2005).
Identification of emotional intonation evaluated by fMRI. Neuroimage, 24(4), 1233-1241.
Winkworth, A. L., Davis, P. J., Adams, R. D., & Ellis, E. (1995). Breathing patterns during
spontaneous speech. Journal of Speech and Hearing Research, 38(1), 124-144.
Yao, B., Belin, P., & Scheepers, C. (under review). Brain „talks over‟ boring quotes: Top-down
activation of voice-selective areas while listening to monotonous direct speech quotations.
NeuroImage.
Yao, B., Belin, P., & Scheepers, C. (2011). Silent reading of direct versus indirect speech activates
voice-selective areas in the auditory cortex. Journal of Cognitive Neuroscience, 23(10),
3146-3152.
Yao, B., & Scheepers, C. (2011). Contextual modulation of reading rate for direct versus indirect
speech quotations. Cognition, 121(3), 447-453.
Yoo, S. S., Lee, C. U., & Choi, B. G. (2001). Human brain mapping of auditory imagery: Eventrelated functional MRI study. Neuroreport, 12(14), 3045-3049.
Yuasa, M., Saito, K., & Mukawa, N. (2011). Brain activity associated with graphic emoticons. The
effect of abstract faces in communication over a computer network. Electrical Engineering
in Japan, 177(3), 36-45.
Zatorre, R. J., Evans, A. C., & Meyer, E. (1994). Neural mechanisms underlying melodic
perception and memory for pitch. Journal of Neuroscience, 14(4), 1908-1919.
Zatorre, R. J., Evans, A. C., Meyer, E., & Gjedde, A. (1992). Lateralization of phonetic and pitch
discrimination in speech processing. Science, 256(5058), 846-849.
Zwaan, R. A. (2004). The immersed experiencer: Toward an embodied theory of language
comprehension. Psychology of Learning and Motivation: Advances in Research and
Theory, 44, 35-62.
Zwaan, R. A., Madden, C. J., Yaxley, R. H., & Aveyard, M. E. (2004). Moving words: dynamic
representations in language comprehension. Cognitive Science, 28(4), 611-619.
Zwaan, R. A., Stanfield, R. A., & Yaxley, R. H. (2002). Language comprehenders mentally
represent the shapes of objects. Psychological Science, 13(2), 168-171.
Zwaan, R. A., & Taylor, L. J. (2006). Seeing, acting, understanding: Motor resonance in language
comprehension. Journal of Experimental Psychology: General, 135(1), 1-11.
Page | 200